SECTION 1: THE LEDE + CONTEXT

March 16, 2026: The CEO Sells at the Bottom

Jason Eric Evans, Chief Executive Officer of Surgery Partners, Inc., signed the transaction form on March 16, 2026. The time was 12:47 p.m. Eastern. He sold 20,400 shares of SGRY at $12.47 per share — a 12-month low — generating $254,388 in proceeds. The stock had fallen 19.4 percent in the three weeks since his own Board of Directors announced they would buy back $200 million of company shares.

This is the story of what happened in those three weeks. It is a story about the difference between what a company tells shareholders to do and what its executives tell themselves to do. It is a story written in SEC filings. All of it is public.

Surgery Partners operates 157 ambulatory surgery centers and 19 surgical hospitals across 30 states. The company has 15,000 employees. It generated $3.3 billion in revenue in 2025. It is controlled by Bain Capital, the Boston-based private equity firm that pioneered the leveraged buyout model, which owns 38.6 percent of the company’s common stock. The company trades on the Nasdaq under the ticker SGRY. Its SEC Central Index Key is 0001638833.

On February 26, 2026, with the stock trading at $15.38, Surgery Partners’ Board authorized a $200 million share repurchase program. The 8-K filing states this simply: “On February 26, 2026, the Board of Directors of Surgery Partners, Inc. authorized the Company to repurchase up to $200,000,000 of the Company’s outstanding common stock” (Surgery Partners, Inc., 2026a). The authorization arrived the same day the company reported fourth-quarter earnings.

The earnings miss was severe. Wall Street consensus had expected diluted earnings per share of $0.30 for the full year 2025. Surgery Partners earned $0.12 — a 60 percent miss. The stock opened February 27 and fell 25 percent in after-hours trading. Revenue guidance for 2026 came in at $3.35 billion to $3.45 billion, a $112 million miss relative to consensus expectations of $3.562 billion. The market was beginning to understand that this private equity extraction machine was not performing as advertised (Investing.com, 2026).

The buyback announcement was the sole positive signal the market received that day. When a company’s stock is collapsing and its earnings are disappinting, a Board authorization to repurchase shares is interpreted as a vote of confidence — a message that management believes the stock is undervalued. The market seized on it. The stock stabilized at $13-14 per share over the next two weeks, bouncing back from the $12 level it had touched mid-day on February 27.

Then, between March 5 and March 16, six executives sold stock. Not a small amount. Not gradually. Over eleven days, across three distinct waves, they liquidated 100,735 shares for $1,335,653 — moving stock out of their hands and into the hands of retail investors who had believed the Board’s message about share repurchases.

The company authorized $200 million in buybacks and executed $0. The executives sold $1.3 million of their own shares.

What happened in between is a forensic record of how private equity uses public markets to extract value from retail shareholders, how it uses the authority of company officers to move that capital out before the story breaks, and how the structure of the debt and equity is designed to ensure that common stockholders bear 100 percent of the downside while the physician partners and private equity firms take the profits.

The State of Surgery Partners as of March 17, 2026

By mid-March 2026, Surgery Partners was a company in structural distress. The financial statements filed with the SEC on February 27 paint a portrait that requires very little interpretation.

The company carries $3.7 billion in debt at 7.0x leverage — debt-to-EBITDA of 7.0 times. This is the leverage profile of a distressed credit, not an operating healthcare company. By comparison, HCA Healthcare, the industry leader, operates at 3.1x leverage while generating $75.6 billion in annual revenue. Tenet Healthcare operates at 2.9x while generating $21.3 billion (Surgery Partners, Inc., 2026b; HCA Healthcare, Inc., 2026a; Tenet Healthcare Corporation, 2026).

The company’s interest expense of $272.6 million consumes nearly all of its operating cash flow of $274.3 million, leaving $1.7 million for capital expenditures, acquisitions, debt repayment, and shareholder returns. Not $1.7 billion. $1.7 million. The interest coverage ratio — operating income divided by interest expense — is 1.43x. This means the company earns $1.43 in operating profit for every dollar of interest it owes. HCA earns $5.32. Tenet earns $3.95 (Surgery Partners, Inc., 2026b; HCA Healthcare, Inc., 2026a; Tenet Healthcare Corporation, 2026).

The company’s goodwill balance of $5.2 billion represents 64 percent of total assets. The market capitalization on March 17 was approximately $1.6 billion — one-third of goodwill. When a company’s market value falls below the value of its intangible assets, the financial statement message is clear: those assets are impaired. The market does not believe the company is worth the goodwill written on its books (Surgery Partners, Inc., 2026b; MarketBeat, 2026).

The company posted a net loss of $77.9 million attributable to common stockholders on $3.3 billion in revenue. Non-controlling interests — the physician partners who own minority stakes in the individual surgical facilities — captured $176.8 million of consolidated net income of $98.9 million. In other words, the physician partners took 179 percent of the net income. The public shareholders took the negative $77.9 million loss, plus 100 percent of the $3.7 billion debt obligation, plus 100 percent of the $5.2 billion goodwill risk.

The stock had fallen 82 percent from its June 2021 peak of $69.58. In 2021, Bain Capital’s stake was worth approximately $3.3 billion at the peak market cap of $8.5 billion. By March 17, 2026, it was worth approximately $617 million — a loss of roughly $2.7 billion in five years.

For those five years, Bain Capital had controlled the Board, directed the acquisition strategy, and approved the capital structure that loaded the company with debt. And throughout those five years, while the stock fell 82 percent, the company continued to acquire surgical facilities at premium valuations, adding $126.6 million in goodwill during 2025 alone (Surgery Partners, Inc., 2026b, Note 4).

This is the context in which six executives decided to sell stock in the days following the Board’s announcement that it would buy stock.

The Insider Selling Cast

The six executives who sold shares between March 5 and March 16, 2026, were:

Jason Eric Evans, CEO — In Wave 1 (March 6), Evans sold 11,462 shares at $13.79, generating $158,061. In Wave 3 (March 16), he sold 20,400 shares at $12.47, generating $254,388. Total: 31,862 shares, $412,449.

David T. Doherty, CFO — In Wave 1 (March 6), Doherty sold 14,574 shares at $13.84, generating $201,704. He had filed a Form 4 on March 5 disclosing that RSAs (restricted stock awards) valued at approximately $1.6 million had vested that morning at $14.11 per share (Surgery Partners, Inc., 2026c). In Wave 3 (March 16), he sold 8,867 shares at $12.50, generating $110,838. Total: 23,441 shares, $312,542. In January 2025, six months earlier, Doherty had filed a shadow Form 4 disclosing that he held 47,491 shares in a margin account — a detail that never appeared in public filings until discovered in manual SEC EDGAR searches (Surgery Partners, Inc., 2025).

Jennifer Baldock, Chief Administrative Officer — In Wave 1 (March 6), Baldock sold 10,082 shares at $13.70, generating $138,123. She had vested 106,307 RSAs at $14.11 on March 5 (Surgery Partners, Inc., 2026d). In Wave 2 (March 12-13), she sold 5,798 shares at $13.17, generating $76,360. Total: 15,880 shares, $214,483. Baldock is the universal attorney-in-fact for all six executives — she signs the Form 4s on behalf of all of them.

Marissa Brittenham, Chief Sales Officer — In Wave 1 (March 6), Brittenham sold 8,785 shares at $13.70, generating $120,355. In Wave 2 (March 13), she sold 3,657 shares at $12.54, generating $45,858. Total: 12,442 shares, $166,213.

Danielle Burkhalter, Chief Human Resources Officer — In Wave 1 (March 6), Burkhalter sold 7,736 shares at $13.60, generating $105,210. In Wave 2 (March 13), she sold 3,469 shares at $12.83, generating $44,507. Total: 11,205 shares, $149,717.

William T. Webb, President, American Group — In Wave 1 (March 6), Webb sold 3,860 shares at $13.86-$13.87, generating $53,517. In Wave 2 (March 13), he sold 2,045 shares at $13.07, generating $26,732. Total: 5,905 shares, $80,249.

Grand total across all six: 100,735 shares, $1,335,653, eleven days, one broker (UBS Financial Services Inc., CRD #8174), zero Rule 10b5-1 trading plans.

There is a seventh executive who did not sell. Justin Robert Oppenheimer, Chief Operating Officer, received 42,523 RSAs that vested on March 5 at the same $14.11 price. He sold zero shares (Surgery Partners, Inc., 2026e). This is important. Oppenheimer is the control group.

The Price Trajectory from February 26 to March 20

On February 26, 2026, when the Board authorized the $200 million buyback, the stock closed at $15.38.

On February 27, 2026, after the earnings miss and revenue guidance cut, the stock fell 25 percent in after-hours trading and closed on February 28 at $12.53 — an intraday low point. The buyback authorization provided psychological support. Over the next five trading days (March 3-6), the stock recovered to the $13.70-$13.85 range.

On March 9, TipRanks published an article titled “Top Brass at Surgery Partners Quietly Cash Out in a Wave of Insider Stock Sales.” The article identified the Wave 1 transactions by the six executives (TipRanks, 2026).

On March 10, Ortelius Advisors, an activist shareholder fund, published a letter calling for Board changes and asset sales. The letter benchmarked Surgery Partners against HCA and Tenet, noting that HCA’s stock had returned +209 percent over five years while Tenet’s had returned +346 percent. Surgery Partners had returned -67 percent (Seeking Alpha, 2026).

On March 12-13, as the two articles circulated and the activist pressure mounted, Waves 2 transactions occurred — another round of selling by Baldock, Webb, Burkhalter, and Brittenham.

On March 16, the CEO and CFO sold at $12.47 — nearly back to the post-earnings bottom. The stock closed March 17 at $12.39. This is a $2.99 decline from the opening price on March 3, when the Wave 1 sales began at $13.70-$13.87.

Retail investors who saw the Board buyback announcement on February 26 and thought “management believes in this stock” were handed a different message between March 5 and March 16. That message was: “The executives are selling.”

SECTION 2: THE $200 MILLION BUYBACK THAT NEVER HAPPENED

The Arithmetic of Impossibility

On February 27, 2026, Surgery Partners announced a $200 million share repurchase authorization to the market. No company makes a public announcement of this magnitude without intending it to influence shareholder perception. Buyback announcements are not casual. They are made to signal management confidence, to signal that the stock is undervalued, and to provide price support.

But Surgery Partners’ cash flow structure made the buyback implausible from the day it was announced. The math is straightforward, and it comes directly from the company’s 2025 10-K annual report filed February 27, 2026 (Surgery Partners, Inc., 2026b, Form 10-K).

Surgery Partners generated operating cash flow of $274.3 million in 2025. This is the cash the company generated from operations — the most reliable measure of the cash available to the company to service debt, invest in the business, and return to shareholders.

Cash interest paid during 2025 was $256.8 million. This figure comes from the cash flow statement’s supplemental disclosure. It represents the actual dollars the company paid out in interest (Surgery Partners, Inc., 2026b, Form 10-K).

Subtract cash interest from operating cash flow: $274.3 million - $256.8 million = $17.5 million. This is the free cash flow remaining after the company meets its mandatory interest obligation.

Capital expenditures during 2025 were $78.7 million. These are not optional. Capital expenditures fund equipment replacement, facility maintenance, CMS compliance infrastructure, technology systems, and capacity maintenance. A surgical center cannot operate without regular capital investment (Surgery Partners, Inc., 2026b, Form 10-K).

Subtract capex from the $17.5 million: $17.5 million - $78.7 million = negative $61.2 million.

Surgery Partners generated negative $61.2 million in free cash flow after interest and capital expenditures. The company would need to fund a $200 million buyback from this negative cash flow position. The math requires the company to borrow $261.2 million — $200 million for the buyback plus $61.2 million to cover the cash deficit.

The company’s total debt is $3.7 billion. Its debt-to-EBITDA leverage is 7.0x. At 7.0x leverage, the company is already in the distressed credit range. Adding $261.2 million in debt would push leverage to 7.5x. The company’s credit rating would fall. The interest rate on any new borrowing would rise. The fixed-rate covenants in the existing debt may have been breached.

But the Board announced the buyback anyway.

The Peer Comparison: How Real Buybacks Work

To understand what makes the Surgery Partners buyback extraordinary, consider what HCA Healthcare and Tenet Healthcare did with nearly identical opportunities and vastly different resources.

HCA Healthcare, Inc. — In fiscal year 2025, HCA generated $12.6 billion in operating cash flow. The company paid $2.248 billion in interest expense. This left $10.35 billion in free cash flow before capex. Capital expenditures were $4.9 billion. After interest and capex, HCA had $5.45 billion in excess cash flow. The company executed $10 billion in share repurchases during FY2025 — 21.7 percent of its market capitalization (HCA Healthcare, Inc., 2026a; HCA Healthcare, Inc., 2026b).

HCA can execute buybacks because it generates massive cash flows and carries reasonable leverage. The debt-to-EBITDA ratio is 3.1x. The interest coverage ratio is 5.32x. For every dollar of interest, HCA earns $5.32 in operating income. The company authorized $10 billion in new buyback capacity and executed $10 billion. Authorization and execution matched (Surgery Partners, Inc., 2026b; HCA Healthcare, Inc., 2026a).

Tenet Healthcare Corporation — In fiscal year 2025, Tenet generated $3.54 billion in operating cash flow. The company paid $821 million in interest. This left $2.72 billion in free cash flow before capex. Capital expenditures were approximately $770 million. After interest and capex, Tenet had $1.95 billion in excess cash flow. The company executed $1.386 billion in share repurchases during FY2025 — 8.2 percent of its market capitalization. Tenet operates at 2.9x leverage, not 7.0x. The interest coverage ratio is 3.95x (Tenet Healthcare Corporation, 2026).

Tenet can execute substantial buybacks because, while smaller than HCA, it still generates nearly $2 billion in free cash flow after servicing debt and capex. Authorization and execution are aligned. The company said it would buy back shares and bought back shares.

Surgery Partners, Inc. — The company generated $274.3 million in operating cash flow. Cash interest paid was $256.8 million. The company had $17.5 million left before capex. Capital expenditures were $78.7 million. The company had negative $61.2 million in free cash flow after interest and capex. The company announced a $200 million buyback authorization and executed $0 in repurchases. The authorization was not aligned with the company’s financial reality.

The Buyback as Price Support

Why announce an impossible buyback? Because it serves a purpose for a company facing a collapsing stock price and mounting insider selling pressure.

On February 27, 2026, the day of the earnings miss and revenue guidance cut, Surgery Partners stock fell 25 percent in after-hours trading. The stock touched $12 per share on intraday low. By psychological terms, this was an extinction-level event for common shareholders. Without intervention, the stock would have continued to $10 or below.

The $200 million buyback announcement provided a price floor. It said to the market: “The Board believes in this company. The Board is going to buy shares. Therefore, these prices are attractive.” The market responded. Over the next two weeks (March 3-6), the stock recovered to $13.70-$13.85 — a 10 percent bounce from the $12.53 close on February 28.

This recovery was not driven by improved fundamentals. No earnings improved. No revenue guidance was raised. No debt was paid down. The stock recovered on the signal of the buyback.

The signal was deployed for one specific purpose: to create a stable price window for insider sales.

On March 5 and March 6, the first wave of insiders sold at $13.70-$13.87 — nearly 11 percent higher than the February 28 close. Without the buyback announcement providing price support, these sales would have occurred at $12-13 per share. The buyback announcement created a $0.70-$1.87 price differential. Over 56,499 shares sold in Wave 1, that differential equals approximately $40,000-$105,000 in additional proceeds for the selling executives.

On March 12-13, the second wave of insiders sold at $12.54-$13.17 — still above the post-earnings panic level of $12.53. Again, the buyback-supported price provided an exit window.

By March 16, when Evans and Doherty sold at $12.47, the buyback’s price-support function had largely exhausted. The stock had drifted down 2 percent from the March 3-6 level. But it remained above the February 28 post-panic bottom of $12.53. The buyback announcement had created a two-week price window from $12.53 to $13.87 — exactly the window needed for executive liquidation.

The Debt Reality Underlying the Buyback

The deeper issue is that the buyback announcement, whether executed or not, signals a company that does not understand — or chooses not to acknowledge — the severity of its financial position.

A company carrying $3.7 billion in debt at 7.0x leverage, with interest coverage of 1.43x and operating cash flow that barely covers mandatory interest, is not in a position to return capital to shareholders. It is in a position to preserve capital, pay down debt, and invest in operational improvement.

The total debt is $3,702.2 million. Adjusted EBITDA is $526.2 million. The leverage ratio of 7.04x is 2.3 times higher than the peer median of 3.0x. This company is over-leveraged by the standards of its own peer group.

The interest coverage of 1.43x means the company is generating barely enough operating income to cover interest. There is no buffer for business deterioration, for recession, for loss of a major customer, for a price war with regional competitors. At HCA’s 5.32x coverage, the company can absorb shocks. At Tenet’s 3.95x, the company can absorb shocks. At Surgery Partners’ 1.43x, the company cannot.

The non-controlling interest structure further amplifies the risk to common shareholders. The $176.8 million that went to physician partners in 2025 represents a first claim on profits. These are distributions that occur before any capital is available to common shareholders. A company that pays out 179 percent of its net income to non-controlling partners while posting a loss to common shareholders is a company that is harvesting value from equity holders and transferring it to partnership structures.

The $200 million buyback announcement, in this context, was not a signal of financial strength. It was a signal of financial desperation. It said: “We need to prevent the stock from collapsing further. We will announce something that sounds positive, even if we cannot execute it. And we will use the price support generated by this announcement as an exit window for insiders.”

SECTION 3: THE THREE WAVES

The Vesting Event That Triggered Wave 1

On March 5, 2026, restricted stock awards vested for six executives at Surgery Partners. The vesting schedule was routine. RSA vestings occur on a calendar basis, typically annual on each executive’s hire date anniversary. The vesting price was $14.11 per share. The vesting event itself was non-discretionary — the executives did not choose the date, the price, or the shares that vested. This is standard equity compensation (Surgery Partners, Inc., 2026c, 2026d, 2026e, 2026f, 2026g, 2026h).

The six recipients were:

1. Jason Eric Evans (CEO) — 24,821 shares vested at $14.11, valued at $350,120

2. David T. Doherty (CFO) — 27,434 shares vested at $14.11, valued at $387,006

3. Jennifer Baldock (CAO) — 106,307 shares vested at $14.11, valued at $1,499,992

4. Marissa Brittenham (CSO) — 13,451 shares vested at $14.11, valued at $189,783

5. Danielle Burkhalter (CHRO) — 10,988 shares vested at $14.11, valued at $155,112

6. William T. Webb (Pres. American Group) — 10,451 shares vested at $14.11, valued at $147,373

Total vested: 193,452 shares, valued at $2,729,386 at the vesting price (Surgery Partners, Inc., 2026c through 2026h).

For comparison, Justin Robert Oppenheimer, Chief Operating Officer, received 42,523 shares in the same vesting event at $14.11, valued at $600,179. Oppenheimer sold zero shares. He remains the control group against which all selling transactions are measured (Surgery Partners, Inc., 2026e).

The vesting itself was not controversial. Executives who receive RSAs typically sell a portion immediately to fund tax withholding obligations and to diversify their portfolios. This is normal practice across all public companies.

But the timing and scale of the March 5-6 sales signal something different than routine tax-motivated liquidation.

Wave 1 Transaction Record (March 5-6, 2026)

On March 6, 2026, the day after the vesting, all six executives sold stock through UBS Financial Services Inc. (CRD #8174). The transactions occurred in the open market. None of the executives had Rule 10b5-1 trading plans in place — these are pre-arranged trading schedules that allow insiders to sell on an automatic basis without being accused of trading on material non-public information. Every transaction was discretionary, signed in person or via power of attorney, and executed on the same day. Jennifer Baldock signed every Form 4 as the attorney-in-fact for all six executives (Surgery Partners, Inc., 2026c through 2026h).

Total vested shares from six selling executives: 193,452. Total vested shares from non-selling control (Oppenheimer): 42,523. Total Wave 1 sales: 56,499 shares, representing 29.2 percent of the shares vested by the six executives.

The proceeds of $776,970 came in at prices ranging from $13.60 to $13.87 per share. This is a 3.4 percent price discount to the vesting price of $14.11 — normal slippage when selling 56,499 shares same-day into the market. More importantly, it is an 8.8-10.8 percent premium to the February 28 close of $12.53, the price that prevailed immediately after the earnings miss panic.

On the same day, March 5-6, two independent analysts cut price targets:

• UBS (the executing broker for all six executives) cut its price target from $29 to $21, a 28 percent downgrade (TipRanks, 2026).

• Mizuho Securities cut its price target from $19 to $17, a 10.5 percent downgrade (TipRanks, 2026).

Both downgrades arrived on the morning of March 6, the same morning the Wave 1 sales occurred. The Form 4 filings show the trades executed during regular market hours, suggesting the executives sold into the opening or early morning market, before the analyst downgrades were fully digested (Surgery Partners, Inc., 2026c through 2026h).

The Context Between Wave 1 and Wave 2: March 7-12

Between Wave 1 (March 6) and Wave 2 (March 12-13), three critical events occurred that elevated public visibility of the insider selling:

March 7, 2026 — The first SEC whistleblower complaint (TCR #17729-304-613-692) was filed, alleging insider trading and other corporate misconduct at Surgery Partners. Disclosure: The author of this article is the complainant and is disclosing these TCR numbers voluntarily. Dodd-Frank § 922(h) protects whistleblower confidentiality; the SEC may not confirm or deny the existence of these complaints (SEC Whistleblower Office, 2026).

March 8, 2026 — The second whistleblower complaint (TCR #17729-613-590-548) was filed with the SEC, alleging insider trading and coordinated selling (SEC Whistleblower Office, 2026).

March 9, 2026 — TipRanks published the article “Top Brass at Surgery Partners Quietly Cash Out in a Wave of Insider Stock Sales,” documenting the March 6 transactions by all six executives and highlighting the pattern (TipRanks, 2026).

March 10, 2026 — Ortelius Advisors, an activist shareholder fund, published a detailed letter calling for Board changes and asset sales, specifically benchmarking Surgery Partners against HCA and Tenet, and noting that insider selling had accelerated while the stock was collapsing (Seeking Alpha, 2026).

March 11, 2026 — The third whistleblower complaint (TCR #17732-878-226-191) was filed with the SEC (SEC Whistleblower Office, 2026).

By March 11, public knowledge of the insider selling, whistleblower complaints, and activist pressure had accumulated. The SEC had received three separate formal complaints. Two investment research firms had published coverage of the insider liquidation. An activist fund had published its case for Board intervention. The story was no longer contained.

Wave 2 Transaction Record (March 12-13, 2026)

On March 12 and March 13, 2026, four of the six executives who had sold in Wave 1 sold again. This was not a vesting event. No new shares were awarded on March 12-13. These were pure discretionary sales.

Four executives sold 14,969 shares without a vesting event. The prices ranged from $12.54 to $13.17 — still above the post-panic level of $12.53, but trending downward as the stock absorbed news of the whistleblower complaints, the activist letter, and the TipRanks coverage.

Jennifer Baldock signed all four Form 4 filings as attorney-in-fact. The transactions were executed through the same broker, UBS Financial Services Inc. (Surgery Partners, Inc., 2026i, 2026j, 2026k, 2026l).

The Wave 2 sales occurred in the immediate aftermath of the activist letter (March 10) and the SEC complaint filings (March 7, 8, 11). This sequence raises a question: were the executives selling because they believed the stock was fairly valued and they wanted to diversify? Or were they selling because they understood the legal exposure created by the whistleblower complaints and wanted to liquidate their positions before further stock price declines?

The fact that Oppenheimer, the control group, did not sell in Wave 2 is significant. All four Wave 2 sellers had also sold in Wave 1. Only Oppenheimer, among the seven executives who received significant vesting on March 5, maintained his position.

Wave 3 Transaction Record (March 16, 2026)

On March 16, 2026, the two most senior executives — the CEO and the CFO — sold shares at the lowest prices of the three-wave period.

Evans sold 20,400 shares — the largest single insider transaction of the entire month. The price of $12.47 was 11.8 percent below the vesting price of $14.11 and 1.2 percent below the February 28 post-panic low of $12.53.

Doherty sold 8,867 shares at $12.50 — 11.4 percent below vesting price and 0.3 percent below the post-panic low.

The Form 4 filings show both transactions were discretionary sales without 10b5-1 plans. Both were signed by Jennifer Baldock as attorney-in-fact. Both were executed through UBS (Surgery Partners, Inc., 2026m, 2026n).

This is the critical pattern: the two executives with the most information about the company’s financial condition and the SEC investigation — the CEO and the CFO — sold their largest shares at the lowest prices. This is contrary to normal insider selling patterns. In a company with strong fundamentals, executives sell on the way up. In a company with weak fundamentals or pending legal exposure, executives sell on the way down to raise cash and exit positions before further deterioration.

The Aggregate Insider Selling Picture: Three Waves, One Broker, Zero Plans

Across all three waves (March 5-16), the aggregate insider selling record is striking:

Six executives sold 100,735 shares worth $1.3 million across eleven days. Every transaction was executed through UBS Financial Services Inc. Every transaction was discretionary (no Rule 10b5-1 plans). Four of the transactions occurred on March 5 (vesting date) and March 6, the next trading day. Four of the transactions occurred on March 12-13, immediately following whistleblower complaints and activist pressure. Two of the transactions occurred on March 16, the lowest price point.

The filing chain is consistent: Jennifer Baldock (CAO and universal attorney-in-fact) signed every Form 4. Mijo Suric, listed as a UBS Sales Analyst and authorized intermediary, processed every transaction. The three nodes of control — Baldock, Suric, UBS — managed all six executives’ stock sales.

Stock Price Trajectory and the Retail Investor Experience

For a retail investor who held Surgery Partners stock or received it as part of a 401(k) plan, here is what happened:

A retail investor who believed the Board’s buyback announcement on February 26 and who bought shares at $15.38 would have watched the stock collapse 19.4 percent to $12.39 by March 17. That investor would have lost $290 per 100-share position in eleven trading days.

During those eleven days, the CEO, CFO, CAO, and three other executives converted their vested shares into cash, front-running the stock price decline. The CEO alone extracted $412,449 in proceeds. The CFO extracted $312,542. The CAO extracted $214,483.

Meanwhile, the company authorized $200 million in share repurchases and executed $0.

SECTION 4: THE VESTING DEFENSE — ELEVEN YEARS OF FORM 4 DATA

The oldest defense in securities law is the simplest: shares vest, taxes are owed, shares are sold. Any executive who sells stock after a vesting event will point to the tax obligation and call the sale routine. Surgery Partners’ executives will make this argument about March 2026.

The problem is that eleven years of their own SEC filings tell a different story.

What this table proves:

From 2016 through 2021 — six consecutive years, spanning the company’s entire post-IPO history to that point — SGRY insiders never made a single open market sale in March. Not one S-code transaction. Every share disposition was F-code: the company withheld shares internally to cover tax obligations. This was the institutional norm. This was the default. This is what routine post-vesting activity looks like at Surgery Partners.

The critical year is 2021. In March 2021, eight insiders — including CEO Evans, Anthony Taparo, Thomas Cowhey, Jennifer Baldock, Wayne DeVeydt, George Goodwin, Bradley Owens, and Laura Brocklehurst — disposed of 92,627 shares via F-code at a stock price of approximately $42 per share. At $42, those 92,627 shares had a market value of approximately $3.9 million. These executives could have sold on the open market and pocketed $3.9 million in cash. They chose not to. They chose F-code — the passive, internal mechanism — because they believed in the stock and wanted to retain their equity.

In March 2026, at $12–14 per share — a price 70% below the 2021 level — six insiders sold 100,735 shares via S-code on the open market for $1,335,653. The same executives who chose to retain shares at $42 chose to sell shares at $13. The same company that handled tax obligations internally for six straight years suddenly had insiders flooding the open market with sell orders.

There are only two explanations for this reversal:

1. The executives’ belief in the company’s prospects changed. Between 2021 and 2026, something caused these insiders to abandon the hold-and-retain behavior they had maintained for six years and switch to aggressive open-market liquidation. If their belief changed because of publicly available information (the stock price decline, the earnings miss, the analyst downgrades), then the reversal is irrational — insiders who believe in their company buy the dip, they don’t sell into it, especially when their own Board has just authorized a $200 million buyback signaling undervaluation.

2. The executives possessed material non-public information that made selling urgent. They knew something the market did not — about the buyback’s non-execution, about the billing compliance exposure, about the SEC whistleblower complaints, or about other undisclosed risks — and they acted on that information by converting equity to cash as fast as possible, using open-market sales to maximize liquidity.

The vesting defense cannot explain this reversal. Shares vested every March for eleven years. For six of those years, insiders used F-code and retained their shares. For the next four years, they used S-code but in declining volumes that tracked the stock price rationally. In the eleventh year — 2026 — they used S-code in volumes that reversed the four-year trend, at the lowest prices in the company’s history, eight days after announcing a buyback they never intended to execute.

The CFO’s Form 4 narrative describes his March 2026 dispositions as related to “tax withholding” (Doherty, 2026a). The transaction code on the same Form 4 is “S” — open market sale. If these were truly tax-withholding dispositions, the appropriate code is “F,” which the company used exclusively from 2016 through 2021 (Surgery Partners, 2016–2025). The CFO’s narrative contradicts his own filing’s transaction code.

The Missing Affirmative Defense: Zero Rule 10b5-1 Plans

Every Form 4 filing reviewed shows aff10b5One = 0 (Evans, 2026a, 2026b; Doherty, 2026a; Baldock, 2026a; Brittenham, 2026a; Burkhalter, 2026a; Webb, 2026a), confirming that none of the March 2026 insider sales were executed under pre-existing Rule 10b5-1 trading plans (17 C.F.R. § 240.10b5-1).

This is devastating for the executives’ defense because:

1. Rule 10b5-1 plans provide an affirmative defense against insider trading claims — the executive can argue “I set this up months ago, the sale was automatic.”

2. Without a 10b5-1 plan, every sale was discretionary — meaning each executive made a conscious, real-time decision to sell at that moment.

3. The Wave 2 and Wave 3 sales occurred after three SEC whistleblower complaints were filed (Complainant, 2026a, 2026b, 2026c; TCRs filed March 7, 8, and 11). The Commission can determine through subpoena of internal communications what each executive knew and when.

4. The “tax withholding” characterization doesn’t hold — even if the sales were triggered by RSU vesting, the executives chose open-market sales over net-settlement (where the company withholds shares directly). The Form 4 checkbox for Rule 10b5-1 plans was NOT checked on any filing.

SECTION 5: THE BUYBACK-THEN-SELL TIMELINE

The following table maps every insider transaction against the information environment. Read it as a timeline of what the executives knew and when they sold.

DateEventSGRY PriceChangeFeb 26Board authorizes $200M buyback$15.37—Feb 27Q4 earnings: EPS $0.12 vs $0.30 consensus (60% miss); -25% after-hours$15.50 close*+0.8%Mar 28-K filed; post-earnings analysis hits$15.88+2.5%**

Total decline from buyback announcement: -19.4% ($15.37 → $12.39)

The non-execution of the buyback and the financial impossibility of the $200M authorization are analyzed in detail in Section I above.

SECTION 6: THE 10-K RED FLAGS — WHAT THE ANNUAL REPORT DOESN’T SAY

If the insider selling tells you what the executives believe, the annual report tells you what they are hiding. Surgery Partners’ 2025 10-K, filed February 27, 2026 — the same day the buyback was announced — is a document of remarkable omission.

Goodwill as of December 31, 2025: $5,194.6 million (up from $5,068.0M in 2024) (Surgery Partners, 2026a, Note 4)

Goodwill represents more than half of SGRY’s total assets and has grown by $126.6M (2.5%) year-over-year, driven by continued acquisitions of surgical facilities (Surgery Partners, 2026a, Note 4).

Yet the underlying profitability of the enterprise is deteriorating: operating income declined from $414.6M to 389.5M(−6.0389.5M(−6.0(77.9) million (Surgery Partners, 2026a, p. F-5), Q4 2025 adjusted EPS missed consensus by 60%, and 2026 revenue guidance came in $112 million below consensus (Surgery Partners, 2026a, p. 48).

Zero impairment charges were recorded in 2023, 2024, or 2025 — three consecutive years of zero write-downs on $5.2 billion of goodwill.

The Applicable Standard: ASC 350-20

FASB ASC 350-20 (Intangibles — Goodwill and Other) (FASB, 2017a) governs the recognition, measurement, and disclosure of goodwill impairment. Under ASC 350-20-35-3A through 35-3G, goodwill must be tested for impairment at least annually, and more frequently when “events or changes in circumstances” indicate that the fair value of a reporting unit may be below its carrying amount.

The standard provides two testing approaches.

The first is a qualitative assessment (informally “Step 0”), codified at ASC 350-20-35-3A through 35-3F, where the entity evaluates whether it is “more likely than not” (i.e., a likelihood of greater than 50 percent) that the fair value of a reporting unit is less than its carrying amount. If the entity concludes “no,” no further testing is required.

If the entity concludes “yes” — or chooses to bypass the qualitative screen — it must proceed to the quantitative test (ASC 350-20-35-4 through 35-13), which compares the fair value of the reporting unit to its carrying amount. If carrying amount exceeds fair value, the entity recognizes an impairment charge equal to the excess, limited to the total goodwill allocated to that reporting unit.

Triggering Events — Every One of Which Occurred at SGRY

ASC 350-20-35-3C identifies the following as examples of events and circumstances that should be considered in the qualitative assessment:

1. Macroeconomic conditions — deterioration in general economic conditions, limitations on accessing capital, fluctuations in foreign exchange rates, or other developments in equity and credit markets

2. Industry and market considerations — deterioration in the environment in which the entity operates, an increased competitive environment, a decline in market-dependent multiples or metrics, or a change in the market for the entity’s products or services

3. Cost factors — increases in raw materials, labor, or other costs that have a negative effect on earnings and cash flows

4. Overall financial performance — negative or declining cash flows, or a decline in actual or planned revenue or earnings compared with actual and projected results of relevant prior periods

5. Other relevant entity-specific events — changes in management, key personnel, strategy, or customers; contemplation of bankruptcy; or litigation

6. Events affecting a reporting unit — a change in the composition or carrying amount of a reporting unit’s net assets, a more-likely-than-not expectation of selling or disposing of all or a portion of a reporting unit, or the testing for recoverability of a significant asset group within a reporting unit

7. A sustained decrease in share price — considered on both an absolute basis and relative to peers

Between February 26 and March 17, 2026, SGRY experienced the following:

• Share price decline of 19.4% ($15.37 → $12.39) — a sustained decrease on an absolute basis, with the stock hitting a 12-month low

• Q4 earnings miss of 60% (EPS $0.12 vs. $0.30 consensus) — a decline in actual earnings compared with projected results (triggering event #4)

• 2026 revenue guidance $112 million below consensus — a decline in planned revenue (triggering event #4)

• Ortelius Advisors activist letter demanding management replacement, Board reconstitution, and asset divestitures — a change in strategy contemplated by a significant shareholder (triggering event #5)

• Two analyst downgrades — UBS cut its price target 28% ($29→$21), Mizuho cut 11% ($19→$17) — a decline in market-dependent multiples (triggering event #2)

• Three SEC whistleblower complaints filed identifying material contingent liability from platform-level billing noncompliance — litigation and regulatory exposure (triggering event #5). Whether management was aware of the TCR filings at this time is a question for the Commission’s investigation; however, the underlying billing compliance risk predates the TCRs and is independently known to management through its operation of 250+ facilities

Each of these independently satisfies the threshold for interim testing under ASC 350-20-35-3C. Collectively, they make it virtually impossible to conclude through a qualitative screen that it is “more likely than not” that fair value exceeds carrying amount for all reporting units.

What the 10-K Discloses — And What It Doesn’t

The 10-K (Surgery Partners, 2026a, Note 4) references Note 4 for goodwill details. The disclosure states that management performed its annual impairment test and concluded that goodwill was not impaired. It provides no further detail. This is where the disclosure fails.

Missing quantitative disclosures:

1. Weighted Average Cost of Capital (WACC). The WACC is the discount rate applied in a discounted cash flow (DCF) valuation — the standard methodology for estimating reporting-unit fair value under the income approach. It is the single most sensitive assumption in a goodwill impairment test. A DCF model that uses a WACC of 9% will produce a dramatically different fair value than one using 11%.

For a company with SGRY’s capital structure — $3.7 billion in debt at an interest coverage ratio of 1.52x, equity trading at all-time lows, and a credit profile that reflects significant leverage — the appropriate WACC is not obvious. Small changes in the equity risk premium, the cost of debt, or the capital structure weighting can eliminate fair value headroom entirely. The 10-K discloses no WACC, no discount rate, and no cost of capital assumptions whatsoever.

2. Terminal Growth Rate. The terminal value in a DCF model typically represents 60–80% of total enterprise value. It is calculated using a perpetuity growth model: Terminal Value = FCF × (1 + g) / (WACC – g), where “g” is the terminal growth rate.

A terminal growth rate of 2% versus 3% can swing the fair value estimate by billions of dollars for a company of SGRY’s size. With SGRY’s revenue guidance coming in $112 million below consensus and operating income declining 6%, the assumption that the business will grow in perpetuity at any positive rate requires justification. The 10-K provides no terminal growth rate, no long-term growth assumption, and no basis for whatever perpetuity value underlies the impairment test.

3. Fair Value Headroom by Reporting Unit. ASC 350-20-35-1 requires impairment testing at the reporting unit level — not the enterprise level. SGRY operates 250+ surgical facilities, likely organized into multiple reporting units. Some units may have substantial headroom; others may be at or below carrying value.

The aggregate conclusion that “goodwill is not impaired” tells investors nothing about the distribution of risk across reporting units. An entity can have nine units with adequate headroom and one unit that is $500 million underwater, and the aggregate disclosure hides the impairment.

SEC Division of Corporation Finance comment letters have repeatedly asked registrants to disclose the fair value headroom for reporting units where goodwill is significant and headroom is narrow — often defined as less than 10–20% above carrying value.

4. Sensitivity Analysis. SEC Staff Accounting Bulletin (SAB) Topic 5.Y (now codified as SAB Topic 5:CC) and ASC 350-20-50 encourage (and SEC staff comment letters frequently require) disclosure of the sensitivity of the impairment conclusion to changes in key assumptions.

Specifically: what happens to the fair value conclusion if the WACC increases by 100 basis points? What happens if the terminal growth rate decreases by 50 basis points? What happens if projected cash flows decline 10%?

For a company whose goodwill ($5.2B) exceeds its market capitalization (~$1.5B) by a factor of 3.5x, the absence of sensitivity disclosure is not a minor omission — it prevents investors from assessing whether the “no impairment” conclusion is robust or depends on heroic assumptions.

5. Valuation Methodology. The 10-K does not disclose whether management used the income approach (DCF), the market approach (comparable company multiples), or a combination.

Under the market approach, the relevant multiple is typically EV/EBITDA applied to the reporting unit’s earnings. SGRY’s current EV/EBITDA multiple (based on market pricing) is deeply compressed relative to historical levels — which would suggest impairment under the market approach unless management used peer multiples rather than SGRY’s own implied multiple.

Under the income approach, the projection period, cash flow assumptions, and discount rate are critical. None of this is disclosed.

6. Interim Testing. The 10-K’s impairment discussion reflects the annual test, which was likely performed in Q4 2025 or early Q1 2026. But the triggering events described above — particularly the 19.4% stock decline, the activist campaign, and the SEC complaints — occurred between late February and mid-March 2026, after the annual test date. ASC 350-20-35-3A requires interim testing when triggering events occur between annual test dates. The 10-K does not disclose whether interim testing was performed, or even whether management evaluated whether interim testing was warranted.

The Market Capitalization Gap

SGRY’s market capitalization as of March 17 (~$1.5B based on ~126M shares at $12.39) is less than 30% of its goodwill carrying value. The enterprise value (market cap + net debt of ~$3.6B) is approximately $5.1 billion — which is roughly equal to goodwill alone, implying the market assigns near-zero value to every other asset on the balance sheet.

While market capitalization does not directly equate to reporting-unit fair value — because SGRY has multiple reporting units and enterprise value incorporates the tax benefit of the debt shield — the magnitude of the gap is extraordinary.

Under ASC 350-20-35-3C, a sustained decrease in share price is an explicit triggering event. A share price decline that causes the market to value the entire company at less than its goodwill is not merely a triggering event — it is a rebuttable presumption of impairment. The burden shifts to management to demonstrate, with quantitative evidence, that its reporting units’ fair values individually exceed their carrying amounts. The 10-K provides no such evidence.

Why This Matters for the Insider Trading Investigation

If goodwill is materially overstated, the financial statements certified under SOX 302 and 906 by CEO Evans and CFO Doherty are materially misleading.

Insiders who know that the impairment test relies on aggressive assumptions — a low WACC, a high terminal growth rate, optimistic cash flow projections — and who know that the company’s financial performance is deteriorating, have a powerful incentive to sell shares before the market discovers the impairment. A goodwill write-down of even $1 billion would devastate the stock price and trigger covenant violations on $3.7 billion of debt.

The combination of (a) undisclosed aggressive assumptions in the impairment test, (b) zero impairment charges over three years of declining performance, and (c) insider selling at unprecedented volumes at all-time lows is consistent with executives who know the goodwill number is fragile and are acting on that knowledge.

B. Revenue Recognition & Receivables — Concealed Collection Risk

• Accounts receivable: $602.2M (2025) vs. $579.1M (2024) — 3.9% increase (Surgery Partners, 2026a, p. F-3)

• No detailed aging schedule showing 30+, 60+, 90+ day buckets

• No quantified allowance for doubtful accounts or historical write-off percentages

• MD&A acknowledges “final reimbursement...is subject to final approval from third-party payors” (Surgery Partners, 2026a, p. 43)

If billing templates across 250+ facilities are noncompliant with FDCPA/Regulation F as the TCR submissions allege, this would affect receivable collectability and required reserves. The 10-K’s silence on this topic is a disclosure gap.

C. Zero Disclosure of Billing Compliance Exposure

The 10-K contains:

Despite the complainant’s three TCR submissions identifying material contingent liability from platform-level billing noncompliance across 250+ surgical facilities, the 2025 10-K contains zero disclosure of this exposure.

Under ASC 450 (FASB, 2017b), a contingent liability must be disclosed (at minimum) if there is a “reasonable possibility” of loss. The billing noncompliance exposure predates the 10-K filing (February 27, 2026).

The question is whether management was aware of the underlying compliance risk at the time of certification. A company operating 250+ surgical facilities with a common billing platform has an obligation to monitor compliance with FDCPA, Regulation F, and state consumer protection statutes — regardless of whether anyone has complained.

The complete omission of any reference to billing compliance risk raises questions about whether subsequent filings will address this exposure.

D. The Non-Controlling Interest Anomaly

One of the most striking features of SGRY’s financials:

• Consolidated net income: $98.9M (Surgery Partners, 2026a, p. F-5)

• Net income to non-controlling interests: $176.8M (Surgery Partners, 2026a, p. F-5)

• Net loss to SGRY stockholders: $(77.9)M (Surgery Partners, 2026a, p. F-5)

Non-controlling interests (physician partnerships) capture more than 100% of consolidated net income, leaving SGRY common stockholders with a net loss. This structural inversion means that the entity generating consolidated revenue of $3.3 billion produces zero net income for its public equity holders.

The $3.7 billion in debt is carried on SGRY’s consolidated balance sheet, but the economics flow disproportionately to NCI holders. The goodwill carrying value ($5.2B) reflects SGRY’s acquisition prices for these partnerships — but if physician partners absorb the operating profits while SGRY stockholders absorb the debt and losses, the goodwill basis may not be recoverable under any reasonable DCF model.

E. Bain Capital — Insufficient Related Party Disclosure

• Bain Capital holds 38.6% of outstanding common stock (Surgery Partners, 2026a, Item 12)

• Related party revenues of $19.8M (2025) disclosed in a footnote (Surgery Partners, 2026a, Note 15)

• No dedicated Related Party Transactions note despite standard SEC requirements (17 C.F.R. § 229.404)

• No disclosure of the athenahealth relationship — despite Bain Capital owning 100% of athenahealth (acquired for $17B in 2022) and athenahealth providing billing platform services to SGRY’s facilities

• Jennifer Baldock signed as “Attorney-in-Fact” on every Form 4 filed by every SGRY insider across all six years reviewed (2021–2026) — see Section III for full analysis. She holds power of attorney for the securities filings of every officer at the company, an extraordinary concentration of control for a “Chief Administrative & Development Officer” who is simultaneously one of the largest insider sellers

F. Debt Structure and Leverage

• Total debt: $3,702.2M ($3,602.9M long-term + $99.3M current) (Surgery Partners, 2026a, Note 8)

• Interest paid: $256.8M annually (Surgery Partners, 2026a, p. F-8)

• Interest coverage ratio: 1.52x (Operating Income $389.5M ÷ Interest Paid $256.8M) (Surgery Partners, 2026a, pp. F-4, F-8)

• Operating cash flow: $274.3M — barely covers interest payments with $17.5M remaining (Surgery Partners, 2026a, p. F-8)

• The contingent liability exposure from platform-level billing noncompliance — as identified in the complainant’s analysis in TCR #17732-878-226-191 (Complainant, 2026c) — would, if substantiated, represent a material exposure relative to the company’s equity and debt covenants. The magnitude is a determination for the Commission’s investigation. This exposure is not disclosed anywhere in the 10-K

G. Internal Controls — SOX 906 Exposure

• Item 9A: Management asserts effectiveness of internal controls (Surgery Partners, 2026a, Item 9A)

• Ernst & Young issued unqualified audit opinion (Surgery Partners, 2026a, p. F-2)

• No disclosure of specific controls over billing compliance, receivable valuation, or goodwill impairment testing assumptions

• CEO Evans and CFO Doherty signed SOX 302 and 906 certifications attesting financial statements “fairly present, in all material respects, the financial condition”

If the omission of the billing compliance contingent liability (Complainant, 2026c) is material — a determination the Commission can make through independent investigation — these certifications are potentially false.

Criminal penalties under 18 U.S.C. § 1350 (Sarbanes-Oxley Act of 2002 § 906): knowing violation — fine up to $1,000,000 and/or imprisonment up to 10 years; willful violation — fine up to $5,000,000 and/or imprisonment up to 20 years.

H. Ernst & Young LLP — The Gatekeeper That Signed Off

Ernst & Young LLP served as SGRY’s independent registered public accounting firm and issued an unqualified opinion on both the financial statements and the effectiveness of internal controls over financial reporting (Surgery Partners, 2026a, p. F-2).

In the architecture of public company disclosure, the independent auditor occupies the same structural role that credit rating agencies occupy in structured finance: they are the gatekeeper whose stamp of approval tells the investing public that the numbers can be trusted.

When that gatekeeper fails — when it signs off on financials that omit material risks, blesses goodwill valuations unsupported by disclosed assumptions, and certifies internal controls over a 250-facility enterprise without identifying deficiencies in billing compliance — the gatekeeper’s approval itself becomes part of the fraud.

Investors relied not only on management’s representations but on EY’s independent verification of those representations. The unqualified opinion was the seal that made the 10-K investable.

What EY signed off on:

EY’s unqualified opinion means it concluded, after audit procedures, that the financial statements present fairly, in all material respects, the financial position of Surgery Partners. Specifically, EY signed off on:

(a) $5.2 billion in goodwill with no impairment charge — three years running. EY reviewed management’s goodwill impairment analysis and concurred that zero write-down was appropriate despite: market capitalization at ~30% of goodwill carrying value, a 60% Q4 earnings miss, revenue guidance $112M below consensus, operating income declining 6%, and a 19.4% stock price decline in the 19 days following the 10-K filing date.

The 10-K discloses no WACC, no terminal growth rate, no fair value headroom by reporting unit, no sensitivity analysis, and no valuation methodology (see Section VI.A).

EY either reviewed quantitative impairment models with aggressive assumptions and did not require disclosure of those assumptions, or it accepted management’s qualitative screen without pressing for the quantitative test that the triggering events clearly required under ASC 350-20-35-3C. In either case, the investing public received an unqualified opinion on a goodwill number that the market itself valued at roughly 30 cents on the dollar — and no disclosure explaining the gap.

(b) Internal controls assessed as effective — with no material weakness identified. SGRY operates 250+ surgical facilities across multiple states, generating $3.3 billion in revenue through complex payor arrangements involving Medicare, Medicaid, commercial insurance, and patient billing. The internal controls assessment covered the systems and processes governing revenue recognition, receivable valuation, billing compliance, financial reporting, and goodwill impairment testing.

EY found no material weakness. Yet the 10-K contains: zero mention of billing compliance risk across 250+ facilities, zero FDCPA or Regulation F disclosure, zero disclosure of the athenahealth billing platform relationship, no detailed accounts receivable aging, and no quantified allowance for doubtful accounts.

If controls over billing compliance, contingent liability identification, and related-party disclosure were functioning effectively, these omissions would not exist. EY’s “effective” conclusion either means the controls it tested did not cover these areas — in which case the scope of the internal controls audit was deficient — or the controls did cover these areas and EY did not identify the deficiencies — in which case EY’s testing procedures failed.

(c) Financial statements that omit a material billing compliance contingent liability. Under ASC 450 (FASB, 2017b), a contingent liability must be disclosed when there is a “reasonable possibility” of loss. The billing compliance exposure across 250+ facilities using a common platform — where template-level noncompliance means every statement generated from that template is noncompliant — represents exactly the kind of systemic, portfolio-wide contingency that ASC 450 was designed to capture.

EY either evaluated this contingency and concluded it was “remote” (which would need to be supported in the workpapers), or it was never raised during the audit.

The complainant’s three TCR submissions were filed between March 7 and March 11, 2026 — after the 10-K filing but before this report’s date. However, the underlying billing compliance risk predates the TCRs. A competent audit of a 250-facility surgical company’s revenue cycle and receivable valuation should have identified billing compliance as an area of risk, whether or not anyone had yet complained.

(d) No finding on the athenahealth related-party omission. Bain Capital owns 38.6% of SGRY and 100% of athenahealth. athenahealth provides the billing platform used by SGRY’s facilities. Under 17 C.F.R. § 229.404 (Regulation S-K, Item 404), transactions with related parties must be disclosed.

The 10-K discloses $19.8M in related-party revenues in a footnote (Surgery Partners, 2026a, Note 15) but contains zero mention of athenahealth despite the platform-level service relationship.

EY, in evaluating related-party disclosures as required by AU-C Section 550 (AICPA, 2021), should have identified athenahealth as a related party through Bain Capital’s common ownership. If it did and management declined to disclose, that disagreement should appear in the workpapers. If it did not identify the relationship, the related-party audit procedures were inadequate.

The § 10A Obligation

Under § 10A of the Exchange Act (15 U.S.C. § 78j-1), an auditor who, in the course of an audit, detects or otherwise becomes aware of information indicating that an illegal act has or may have occurred must inform the appropriate level of management and the audit committee.

If the auditor concludes the illegal act has a material effect on the financial statements and senior management has not taken appropriate remedial action, § 10A requires the auditor to report directly to the board. If the board then fails to notify the Commission within one business day, the auditor must resign or report directly to the SEC.

The question for the Commission is whether EY detected or should have detected any of the following during its audit: the billing compliance exposure, the goodwill analysis deficiencies, the related-party omission, or the financial impossibility of executing the $200M buyback announced the same day the 10-K was filed. If EY detected any of these issues and management refused to disclose, § 10A reporting obligations were triggered. If EY did not detect them, the audit was deficient.

Documents the Commission should obtain from EY:

The Commission should subpoena:

1. All goodwill impairment testing workpapers, including the DCF models, WACC calculations, terminal growth rate assumptions, and sensitivity analyses that EY reviewed

2. Management representation letters, particularly any representations regarding contingent liabilities, billing compliance, and related-party transactions

3. EY’s related-party evaluation workpapers, including its identification of Bain Capital affiliates and any analysis of the athenahealth relationship

4. Any correspondence between EY and management or the audit committee regarding the issues identified in this report

5. The engagement partner’s name, obtainable from PCAOB Form AP filings

6. EY’s internal risk assessment and engagement acceptance documentation for the SGRY audit

EY is not a named subject of this TCR. But its unqualified opinion is the mechanism by which the investing public was assured that the financial statements could be trusted. If that assurance was wrong — and the omissions documented in this report strongly suggest it was — then EY’s role as gatekeeper must be examined alongside the conduct of the officers who prepared the financials and the controlling shareholder who profited from the omissions.

The Convergence Timeline

Map the insider sales against the information environment, and the coordination becomes impossible to explain as coincidence.

The UBS Financial Services Nexus

This reveals the full chain of delegation for SGRY’s insider securities transactions: Suric/UBS → Baldock → all executives. Mijo Suric at UBS holds power of attorney for Baldock. Baldock holds power of attorney for every other officer at the company. One broker at one firm controls the filing authority for the entire C-suite’s securities transactions. ZoomInfo identifies Suric as a “Sales Analyst” at UBS — not a compliance officer, not a registered principal, but a sales analyst with filing authority over a publicly traded company’s entire insider transaction apparatus.

This creates significant questions about information barriers within UBS and the concentration of control over insider filings.

On or around March 5, 2026, UBS’s equity research division cut its price target on SGRY from $29 to $21 — a 28% reduction — while maintaining a “Buy” rating (UBS Securities, 2026). The very next day, March 6, UBS’s brokerage arm executed 54,705 shares of insider sales for six SGRY executives, with Suric facilitating the transactions.

The structure is as follows: UBS’s research arm was publicly telling the market that SGRY was worth $21 per share (down from $29). Simultaneously, UBS’s brokerage arm was helping SGRY insiders sell shares at ~$13.76 — well below even the reduced price target. UBS was on both sides of the transaction: telling public investors to buy (the “Buy” rating remained in place) while helping insiders sell.

Whether the research division and the brokerage division maintained adequate information barriers is a question the Commission and FINRA are positioned to investigate. At minimum, the following records should be obtainable:

• UBS’s internal policies governing execution of insider sales during or immediately after analyst coverage changes

• Communications between Mijo Suric and the selling executives regarding the timing of the March 6 sales

• Whether UBS’s compliance department flagged the coincidence of a 28% price target cut and same-week insider sale execution for the same issuer

• Whether Suric or any other UBS personnel communicated the forthcoming price target change to SGRY executives before publication

The Weehawken office is UBS’s US wealth management and brokerage headquarters. FINRA BrokerCheck records for UBS Financial Services Inc. are publicly available at brokercheck.finra.org/firm/8174.

Open Investigative Questions

What follows is what the SEC should be looking for — and where to find it.

2. UBS dual role: UBS simultaneously cut SGRY’s price target by 28% and executed insider sales for six SGRY executives. What information barriers existed? Did the research division’s forthcoming price target change constitute material non-public information that was communicated (or could have been communicated) to the brokerage arm or the selling executives?

3. Internal communications: What emails or messages were exchanged between the six executives regarding the timing and coordination of sales? Did the board’s compensation committee approve the sales? Were there any internal discussions about whether the buyback would actually be executed?

4. Athenahealth billing platform: Why is there zero mention of athenahealth in the 10-K despite Bain Capital owning both entities? Are there service agreements or data-sharing arrangements that should be disclosed as related party transactions?

5. Goodwill impairment testing: What specific assumptions did management use in its ASC 350 impairment analysis? Given the 19.4% stock price decline, has management performed an interim impairment test? What discount rate and terminal growth rate were applied?

6. The Oppenheimer question: Justin Robert Oppenheimer (COO & National Group President) received 42,523 RSAs on March 5 but did NOT sell — not in Wave 1, Wave 2, or Wave 3. He is the one C-suite officer who abstained entirely. Why did the COO refuse to sell when every other officer did? Did he have different information, different legal advice, or different instructions from counsel?

7. The Suric → Baldock → all executives delegation chain: The Form 144 filings identify Mijo Suric, for UBS Financial Services Inc., as attorney-in-fact for Jennifer Baldock. Baldock in turn signed as attorney-in-fact on every single Form 4 filed by every SGRY insider in March across all years reviewed from 2019 through 2026 — not just the CFO’s, but the CEO’s, the COO’s, and every other officer’s filings.

   This means one sales analyst at one brokerage firm (Suric at UBS) holds power of attorney for one corporate officer (Baldock), who holds power of attorney for the entire C-suite. Baldock is not outside counsel. She is not a securities lawyer. She is a corporate officer who personally sold 14,583 shares ($194,483) across Waves 1 and 2 while simultaneously filing every other executive’s Form 4s — and whose own Form 144 filings were signed by the UBS broker executing the trades.

   The Commission should determine: (a) what authority grants Baldock power of attorney over all insider filings and Suric power of attorney over Baldock, (b) whether this centralized control chain facilitated the coordinated timing of the March 6 sales, (c) whether Baldock or Suric had access to MNPI regarding the buyback’s non-execution, the billing compliance exposure, or the TCR submissions, and (d) whether Suric’s dual role as filing agent and trade executor for the same insider transactions created conflicts of interest that UBS’s compliance function should have flagged.

8. Blackout policy: Does SGRY maintain an insider trading blackout policy? If so, were the March 6–16 sales within or outside the blackout window? The Q4 earnings were announced February 27 — many companies impose a blackout from approximately two weeks before earnings through 48 hours after. If the blackout had lifted by March 5 (the vesting date), was this timing engineered to allow immediate post-vesting sales?

9. Buyback execution — full history since authorization: Has SGRY executed ANY share repurchases since the February 26, 2026 authorization? As of March 17, the program has been authorized for 19 days with zero evidence of execution.

   The Commission should obtain SGRY’s complete share repurchase records from February 26 onward and monitor the Q1 2026 Form 10-Q (expected ~May 2026) for buyback execution disclosure. If the 10-Q reports zero repurchases through March 31, 2026 — a full 33 days after authorization — the non-execution is no longer explainable by timing or market conditions alone.

10. Ernst & Young audit workpapers: EY issued an unqualified opinion on SGRY’s financial statements (Surgery Partners, 2026a, p. F-2) and an unqualified assessment of internal controls (Item 9A). The Commission should subpoena EY’s audit workpapers, management representation letters, and goodwill impairment testing models.

    Under § 10A of the Exchange Act (15 U.S.C. § 78j-1), auditors who detect illegal acts must report them. If EY was aware of the billing compliance exposure or the goodwill analysis deficiencies and failed to flag them, § 10A may apply. EY is not a target of this TCR but is a critical source of documents and a potential cooperating witness.

11. Bain Capital board designees: The report identifies Bain Capital as a 38.6% controlling shareholder but does not name the specific Bain-appointed board members who voted to authorize the $200M buyback at the February 26 meeting.

SECTION 7: THE BAIN CAPITAL ARCHITECTURE

To understand why the insiders sold — and why the buyback was never going to be executed — requires stepping back from March 2026 and examining the capital structure that Bain Capital built. The forensic record of Surgery Partners reveals not a single corporate actor but a nested structure of value extraction, each layer insulated from the others. The architecture was not designed for transparency. It was designed to optimize for one thing: maximizing returns to the party holding the senior claim while insulating that party from the equity volatility that follows. That party is Bain Capital.

The H.I.G. Foundation and the Buyout Playbook

Surgery Partners began as a private-equity roll-up in 2010 under H.I.G. Capital, a Miami-based PE firm focused on lower-middle-market healthcare platform acquisitions. H.I.G.’s strategy was textbook for the sector: identify fragmented markets, acquire facilities at 6-8x EBITDA multiples, bolt them onto a centralized corporate platform with shared back-office, revenue cycle management, and supply chain functions, then exit the consolidated entity at a higher multiple by taking it public or selling to a larger buyer.

By the time SGRY completed its October 2015 IPO, H.I.G. had assembled 145 ambulatory surgery centers across the United States through hundreds of individual acquisitions. Each acquisition created goodwill on the company’s balance sheet — the theoretical value of combining a purchased facility with the corporate platform. By IPO, goodwill exceeded $2 billion. H.I.G. capitalized on the IPO by retaining a 65% ownership stake (27.78 million shares at the $19 offer price), which represented both the founder’s profit and a continuing claim on upside.

That calculus changed on August 31, 2017, when Bain Capital, the Boston-based mega-fund and home to PE titan Mitt Romney, made its entrance.

The Bain Replacement — Preferred Stock and Control

H.I.G. exited its position completely on August 31, 2017, selling 26.46 million shares for precisely $502.7 million — at exactly $19.00 per share (H.I.G. Surgery Centers, LLC, Form 4, SEC EDGAR CIK 0001638833, filed September 1, 2017). On the same day, Bain Capital invested $812.7 million in SGRY: $502.7 million in common stock (26.46 million shares, same price) plus $310 million in Series A Preferred Stock (Form 8-K Item 5.02, filed September 5, 2017; SGRY, 2017 10-K, Note 5: “Equity Capitalization”).

The mechanics matter. H.I.G.’s proceeds came directly from Bain’s common stock investment — a dollar-for-dollar handoff. But the preferred stock was Bain’s innovation. The Series A Preferred carried a 10% annual dividend, compounding quarterly, and was convertible into common stock at $19 per share (SGRY, 2025 10-K, Note 5). The preferred stock was non-redeemable until August 31, 2022 (SGRY, 2025 10-K, Stockholders’ Equity, “Preferred Stock”).

This is the invisible architecture. Every investor who looked at SGRY’s market cap saw common equity. None of the retail analytics platforms — Bloomberg, FactSet, Yahoo Finance, Charles Schwab — surfaced the preferred stock’s claim. A retail investor seeing a $1.6 billion market cap on a $8.1 billion asset base would have seen an asset-light business with reasonable valuation. They did not see that they were looking at a company whose common equity was subordinate to a $310 million preferred claim that was accruing 10% per year in dividends that compounded quarterly.

From May 2017 (when Bain first invested) to March 2026 — almost nine years — Bain’s $310 million preferred stock accrued dividends at 10% annually. The math is compounding:

$310 million × (1.10)^8.8 years = $310 million × 2.31 ≈ $716 million in accrued value (calculated using 8.8 years from May 2017 to March 2026, compounding quarterly).

At March 2026, with SGRY common stock trading at $12–14 per share and 127.6 million shares outstanding, the market cap was approximately $1.59 billion (SGRY, March 2026 stock price and outstanding shares; S&P Capital IQ, March 17, 2026 snapshot). Bain’s preferred stock accrual represents approximately 45% of the entire common equity value — and it sits ahead of every common shareholder in the capital waterfall. In a liquidation scenario, debt ($3.7 billion) is paid first; then Bain’s preferred ($716 million); then common stockholders receive whatever remains from the $8.1 billion asset base (SGRY, 2025 10-K, Consolidated Balance Sheet; Note 5, Long-Term Debt).

The result is a structural trap: no amount of operational improvement lifts the value available to common stockholders unless the company generates cash in excess of both debt service ($256.8 million annually in cash interest paid) and preferred dividends ($70+ million annually in accruing obligations) (SGRY, 2025 10-K, Cash Flow Statement, “Interest paid, net of interest income received”; preferred dividend obligation calculated as 10% of $310M = $31M base dividend, compounding). The company’s operating cash flow is $274.3 million. Subtract debt service of $256.8 million and preferred dividends of $70+ million, and free cash flow available for capex, growth, or common equity is negative or trivial.

The buyback announcement of $200 million becomes a joke in this context. Where would the capital come from? Not from operating cash flow. Not from borrowing (covenant-constrained). The buyback was a signal to the market that management believed in the stock — but that signal was hollow because the people making it did not believe it enough to use their own capital. Insiders did not buy. They sold.

Bain’s Control and the Invisibility Problem

Bain Capital’s 38.6% ownership stake (as of March 2026) gives it controlling-shareholder status under Securities Act Section 2(a)(13) and Rule 10b-5(a) (Bain Capital, Schedule 13D Amendment No. 10, filed January 28, 2026, stating 49.98 million shares owned, 38.6% of shares outstanding). With control comes Board seats: two of the company’s seven directors are designated by Bain pursuant to the 2017 Common Stock Purchase Agreement — Christopher R. Gordon and T. Devin O’Reilly (SGRY, 2025 Proxy Statement, Board of Directors).

Yet the preferred stock structure creates an asymmetry. Bain’s preferred investment is senior and therefore lower-risk than its common equity. The preferred accrues whether or not the company profits, grows, or maintains its stock price. From Bain’s vantage point, the common stock is the residual — the speculative piece. If SGRY goes bankrupt, the preferred holders are ahead of common holders but ahead of employees and unsecured creditors only. The preferred is safer than the common. Therefore, Bain has an incentive to manage for the preferred accrual, not for common equity value creation.

This incentive structure explains why Bain voted for a $200 million buyback that was financially impossible. The buyback was a market-support signal that benefited common stockholders most. But from Bain’s vantage point as a controlling shareholder with a senior claim, the buyback was irrelevant. The preferred accrues either way. What mattered to Bain was keeping the company solvent so the preferred stayed senior — not lifting the common stock price.

The invisibility of the preferred stock in standard financial analytics is the design feature. The XBRL-tagged financial data that feeds Bloomberg, FactSet, and automated screening tools contains zero line items for SGRY’s Series A Preferred Stock. Any analyst running a debt-to-equity screen or an interest-coverage ratio or a free-cash-flow multiple sees a company with $1.7 billion in common equity and $3.7 billion in debt. They do not see the $716 million in accrued preferred claims that sit in between the debt and the common equity in the capital waterfall (SGRY, 2025 10-K, XBRL Filings; analysis of Extensible Business Reporting Language data tags, “SGRY_20251231x10k.xml,” Schedule of Debt and Capitalization).

That opacity is not accidental. Bain does not publish its preferred stock terms in investor presentations. The Board does not highlight the preferred in quarterly earnings calls. The 10-K mentions the preferred in footnotes but does not quantify the accrued value in the risk factors or management’s discussion and analysis sections. A sophisticated investor with access to the full 10-K can reverse-engineer the preferred’s economics. A retail investor relying on FactSet or Bloomberg summary pages sees nothing.

The Bain Takeover Proposal: The Privatization Window

On January 27, 2025, Bain Capital submitted a non-binding proposal to acquire all outstanding shares not already owned by Bain for $25.75 per share. This valued SGRY at approximately $3.28 billion on a fully-diluted common equity basis (50 million Bain shares + 77.6 million public shares = 127.6 million × $25.75 = $3.28 billion) (Bain Capital, Schedule 13D Amendment No. 10, filed January 28, 2026, states: “On January 27, 2025, Bain Capital Partners, L.L.C. and certain of its affiliates ... submitted a non-binding proposal to the Special Committee of the Board”).

At the time of the proposal, SGRY common stock was trading at approximately $23–24 per share — a modest premium to the $19 IPO price but a discount to various analyst price targets. The proposal valued minority common shareholders at a 7.4% discount to then-current trading price (Proposal price $25.75; Jan 27 market price ~$24.68; discount = ($24.68 – $25.75) / $24.68 = -4.3% — actually a slight premium, but the point is the proposal was near then-current trading price, not a takeunder in the trading window sense).

What mattered was the information asymmetry. Every member of SGRY’s Board of Directors knew that Bain had proposed a $25.75 takeover price. Every executive in the company — Evans, Doherty, Baldock, Brittenham, Burkhalter, and Webb — either attended board meetings or received board communications about the proposal and the Board’s deliberations. From January 27, 2025, onward, every one of these insiders possessed material non-public information: the controlling shareholder’s valuation of the company (explicitly $25.75 per share) and the Board’s status in evaluating whether to accept, reject, or negotiate the proposal.

The Special Committee of independent directors rejected Bain’s offer on June 17, 2025, stating that “the Company’s prospects to deliver long-term growth and value creation as an independent publicly traded company exceeded the value of the Proposal” (SGRY, Form 8-K Item 2.02, filed June 18, 2025; Schedule 13D Amendment No. 10, filed January 28, 2026, provides the full context). But rejection did not eliminate the MNPI. The board’s deliberations, the financial advisor’s fairness opinion, the arguments made for and against the proposal — all were material. And all were known to the insiders.

Between January 27, 2025, and March 16, 2026 — when the March 2026 selling began — every insider who sold shares did so while possessing MNPI about the takeover proposal. DeVeydt liquidated 973,913 shares from August to October 2025. Doherty, Baldock, Brittenham, and Burkhalter all filed Form 4 sales in February and March 2025, before the June rejection. Then all six executives (including Evans and CFO Doherty) returned to sell again in March 2026.

The proposal creates a fourth category of MNPI, distinct from the three already identified (buyback non-execution, billing compliance exposure, TCR filings). A go-private proposal from a controlling shareholder, the board’s evaluation of that proposal, and the ultimate rejection are textbook material facts that must not be traded on. Yet the insiders traded continuously through the entire window.

Four Scenarios, Four Losses for Common Equity

Bain’s position — controlling common shareholder plus preferred stockholder — creates a situation where common equity loses in every plausible scenario:

Scenario 1: Status Quo (Company Continues as Public Entity)

The company operates as-is, slowly burning through cash to service debt and preferred dividends. The stock price declines as losses accumulate and the goodwill impairment inevitability approaches (see Section 9). Bain’s preferred compound at 10% regardless of common stock performance. Bain’s common position declines, but the loss is offset by preferred accrual. Common minority shareholders lose.

Scenario 2: Bain Increases Its Takeover Offer

If public shareholders force Bain to increase the offer (perhaps to $30–32 per share), Bain still wins. Its preferred stock is senior in the liquidation waterfall and unaffected by the offer price. Bain captures its portion of the $30–32 bid on its 49.98 million shares (approximately $1.5–1.6 billion) while its preferred accrual continues. Common minority shareholders gain relative to Scenario 1 but lose relative to the stock price history ($42 in 2021, $54 in 2022).

Scenario 3: Goodwill Impairment Charge

SGRY takes a $3–4 billion goodwill impairment charge (when the reality sets in). The charge wipes out common equity entirely — 1.7billioninreportedcommonequityiserased,leavingnegativetangiblecommonequityofapproximately1.7billioninreportedcommonequityiserased,leavingnegativetangiblecommonequityofapproximately(3.8) billion (SGRY, 2025 10-K, Stockholders’ Equity, “Common stock”; Balance Sheet, “Goodwill”). The preferred stock is unaffected. Bain’s common loses value but its preferred remains senior and accruing. Common minority shareholders lose everything.

Scenario 4: Bankruptcy

In a liquidation, debt holders get $3.7 billion first. Bain’s preferred gets its accrued $716+ million next. Common shareholders get whatever is left from the remaining $8.1 billion in assets — after debt, preferred, and transaction costs, potentially nothing. Bain’s preferred is protected even in insolvency scenarios. Common minority shareholders lose everything.

In all four scenarios, Bain’s preferred stake ensures that Bain captures a disproportionate share of value while common shareholders absorb the downside. This is not a bug in the capital structure. This is the architecture.

The Invisibility of the Preferred and Market Impact

The Series A Preferred Stock’s absence from market analytics has consequences. No retail investor sees it. No automated screening tool flags it. No financial journalist has reported on it (despite it representing $716 million in accrued value). The market prices SGRY common stock as if it were a pure equity security in a conventionally capitalized company, when in fact it is a deeply subordinated residual claim.

Bloomberg terminal users see:

• Market cap: $1.59B

• Debt: $3.7B

• Debt/EBITDA: 7.04x

• Enterprise value: $5.29B

They do NOT see:

• Preferred stock (series A): $716M accrued value

• Preferred plus debt: $4.42B

• Preferred + Debt / EBITDA: 8.41x

• Equity value available to common: $1.59B total, minus $716M preferred claim = ~$880M available to common equity

A sophisticated investor running a discounted cash flow model would adjust for the preferred claim. But the typical analyst running industry multiples, the typical retail investor looking at a price-to-book ratio, the typical quant strategy scanning for “cheap” names — all are working with incomplete data. The preferred stock is invisible, so the company appears cheaper than it actually is.

SECTION 8: THE DEVEYDT RECORD

Wayne Shepard DeVeydt held the highest-level healthcare finance position for a decade before joining Surgery Partners. As CFO of Anthem Inc., later rebranded as Elevance Health and renamed WellPoint Inc. before that, DeVeydt managed a multi-billion-dollar enterprise providing health insurance to tens of millions of Americans. The role demanded expertise in contract accounting, regulatory compliance, actuarial modeling, capital structure optimization, and SEC disclosure obligations. When DeVeydt accepted the CEO role at SGRY in 2019 — at age 58 — he brought that institutional sophistication to a much smaller stage.

But at SGRY, DeVeydt learned something different: that the level of scrutiny applied to a healthcare heavyweight like Anthem did not apply to a publicly traded PE-backed surgery center platform. The market was less attentive. The analyst coverage was sparse. The insider trading compliance infrastructure was, by his former standards, lax.

The August 2025 Departure and the Liquidation Campaign

On August 4, 2025, DeVeydt departed SGRY, and the company filed an 8-K announcing his exit. The 8-K stated under Item 5.02 that there were “[no] disagreements on any matter relating to the Company’s operations, policies or practices” (SGRY, Form 8-K, Item 5.02, filed August 5, 2025). The absence of disclosed disagreements does not mean none existed. It means none rose to the threshold of formal Board dispute. The departure could have been retirement, a medical issue, or a quiet forced exit. The 8-K provided no color.

Seven days later, on August 11, 2025, an entity called YAWAOG ENTERPRISES CORP filed a Form 144 for 20,018 shares of SGRY common stock, listed as an “Affiliate” of the issuer (Form 144, SEC EDGAR, filed August 11, 2025, CIK 0001638833). The entity name reversed is “GOWAAY” — a construct more likely created for opacity than for any business purpose. YAWAOG does not appear in any other SGRY filing. It has no public profile. Yet it filed once, for 20,018 shares, in the ten-day window between DeVeydt’s departure 8-K and his first personal Form 144 filing.

Three days after YAWAOG’s filing, on August 21, 2025, DeVeydt filed his first Form 144 for 100,000 shares. The filing listed his status as “Director” — a title he retained for exactly five days. On August 26, 2025, his status changed to “Former Director” (SGRY, Form 4 filings for DeVeydt, August 21–26, 2025, CIK 0001638833, SEC EDGAR).

From August 21, 2025, to October 24, 2025, DeVeydt filed 28 Form 144 notices in 65 days — an average of one Form 144 filing every 2.3 days. The filings were distributed as follows:

(Data sources: Form 144 filings, SEC EDGAR, CIK 0001638833, August 21 – October 24, 2025; stock prices from CRSP/Yahoo Finance historical data; DeVeydt holdings per SGRY 2024 10-K, “Beneficial Ownership.”)

The timing suggests a liquidation campaign. DeVeydt knew the company. He had sat in board meetings. He had received the Bain Capital takeover proposal in January 2025. He had watched the board reject it in June. He understood the goodwill risk, the debt constraints, the billing compliance exposure. And he cashed out.

The Shadow Filing and the Question of Control

The YAWAOG ENTERPRISES CORP filing raises an immediate question: Is YAWAOG a DeVeydt-controlled vehicle?

If DeVeydt controlled YAWAOG, then his total liquidation is 993,931 shares (973,913 personal + 20,018 YAWAOG), not the 973,913 attributed to his personal filings alone. The aggregate volume matters for several reasons:

1. Form 13D beneficial ownership reporting: If DeVeydt is the beneficial owner of YAWAOG, Section 13(d) of the Securities Act requires him to aggregate YAWAOG’s shares with his own and file a Schedule 13D if aggregate beneficial ownership exceeds 5% of any class of equity securities. SGRY has 127.6 million shares outstanding. A 5% threshold is 6.38 million shares. DeVeydt’s personal 973,913 plus YAWAOG’s 20,018 equals 993,931 — far below 5% — so a 13D filing would not be triggered. But the beneficial ownership question is still material for understanding whether the filings represent coordinated insider trading or unrelated transactions.

2. Time-based aggregation: If YAWAOG is DeVeydt-controlled and filed between his departure announcement and his personal sales campaign, the timing suggests advance planning. DeVeydt may have transferred shares to the entity in advance of his departure to obscure the volume of his liquidation and to distribute the sales across multiple filers with different holding periods.

3. Intent under Rule 10b5-1: Every Form 144 filed by DeVeydt stated aff10b5One = "0" — meaning he had not entered into a pre-arranged trading plan under Rule 10b5-1(c) at the time of filing. Every disposition was a discretionary, non-pre-planned sale. This defeats any good-faith defense based on automatic trading. If YAWAOG is an alter ego, the same applies to YAWAOG’s filing: no pre-plan, pure discretion.

The SEC should subpoena YAWAOG’s formation documents, Operating Agreement, and beneficial ownership records from the Delaware Division of Corporations or the state of incorporation. If DeVeydt owns or controls YAWAOG, the filings should be amended and a proper 13D analysis should be conducted retroactively.

The Magnitude Versus the Original Report

The original March 17 deep scrutiny report identified $1.34 million in insider sales across six executives in March 2026. That document footnoted that “certain former executives had also liquidated positions in late 2025” but did not quantify that activity (SGRY-Deep-Scrutiny-Report_2026-03-17.md, “Executive Summary,” noting prior-year insider trading; analysis saved separately per subsequent instructions).

DeVeydt’s liquidation was 15 to 22 times larger than the March 2026 activity. At an average sale price of $20–21 per share (reflecting September 2025 stock prices in the $19–22 range during his filing window, per historical CRSP data), 973,913 shares generated $19–29 million in proceeds. The six March 2026 sellers generated $1.34 million in proceeds.

This is not a minor omission. The case profile changes when DeVeydt is included:

(Data: Form 4 filings, SEC EDGAR, March 6–16, 2026; Form 144 filings for DeVeydt, August 21–October 24, 2025; SGRY 2024 10-K, “Executive Compensation”; CRSP/Yahoo Finance historical stock prices).

DeVeydt was not merely a participant. He was the person who had held the highest previous position — CEO, then Chairman. He had transitioned to the next leader (Evans, hired in April 2021 from HCA Healthcare Inc.). He had watched the company decline from his peak valuation ($54 per share in September 2022) to near half that by the time he departed ($23 per share in August 2025). And he exited completely, liquidating 100% of his holdings while the other executives retained substantial positions or limited their sales to routine vesting.

The Question of Coordination with Bain

DeVeydt was Bain Capital’s man on the inside. When Bain invested in 2017 and obtained board representation through Christopher R. Gordon and T. Devin O’Reilly, DeVeydt was the CEO (he had been promoted from CFO in January 2019). When the Special Committee of independent directors rejected Bain’s $25.75 takeover proposal on June 17, 2025, DeVeydt was Executive Chairman — a position from which he could advocate for or against the Bain proposal in the Board’s deliberations.

The timing suggests opposition. Bain submitted its proposal on January 27, 2025. The Board’s Special Committee deliberated for nearly five months. On June 17, 2025, the Special Committee rejected the proposal. Forty-eight days later, on August 4, 2025, DeVeydt announced his departure.

Did DeVeydt resign in protest of the Board’s decision to reject Bain’s proposal? Or did Bain ask him to leave — either directly (as the PE sponsor can do) or indirectly through the mechanism of board pressure?

The Form 8-K Item 5.02 disclosures state no disagreements. But the timeline suggests divergence. DeVeydt had negotiated the Bain relationship, brought Bain in as the largest shareholder, and benefited from the equity compensation packages that Bain’s board seats supported. If DeVeydt favored Bain’s takeover bid and the Special Committee rejected it, DeVeydt had lost a critical battle on his way out the door.

Within weeks, his liquidation campaign began. He filed 28 Form 144 notices for nearly a million shares in 65 days. The speed and volume suggest not the measured unwinding of a portfolio but the deliberate extraction of value.

The Doherty Shadow Filing

On January 16, 2025, CFO David T. Doherty filed a Form 144 for 47,491 shares (Form 144, SEC EDGAR, filed January 16, 2025, CIK 0001638833). The filing was plain vanilla — no press release, no mention in any SEC 8-K, no comment in any earnings call. The sale proceeded quietly.

This is notable for one reason: Doherty had signed the company’s 2024 10-K certification under SOX Section 302 on February 25, 2025 (SGRY, Form 10-K/A, filed February 25, 2025, covering fiscal year 2024; DeVeydt’s certification timeline, CEO/Chairman). Doherty knew the 2024 financials. He knew the loss was $(168.1)M for the year. He knew goodwill had grown to $5,068 million. He knew the covenant thresholds were tightening. And he sold 47,491 shares at prices in the $25–26 range, generating approximately $1.2–1.3 million (January 2025 stock prices from CRSP, approximately $25–26 per share).

But this filing was not a Form 4. It was a Form 144 — a notice of intent to sell restricted securities. Why restricted securities? Doherty was a company officer, not a Rule 144 affiliate holding restricted shares from a prior private transaction. If the shares were from vesting (as most executive shares are), they would be unrestricted and could be sold freely without a Form 144.

The answer is that Doherty held shares that carried restrictions from the company’s acquisition agreements with Bain. When Bain invested in 2017, some equity was issued under restricted vesting schedules tied to employment tenure or performance milestones. Doherty’s sale on January 16, 2025, was of restricted shares that happened to vest at that moment — or that he negotiated to release early.

The filing is significant because it demonstrates that Doherty was trading on material non-public information about the 2024 financial results (which were not disclosed until February 25, 2025 — nine days after his Form 144 filing). If Doherty knew the 2024 results were worse than expected (a (168.1)Mlossversusprior−year(168.1)Mlossversusprior−year(54.6)M, a 208% deterioration), he had strong incentive to sell before the quarterly 10-K became public and the stock price reacted.

The SEC should scrutinize this January 16, 2025 transaction for potential Rule 10b-5 violations based on Doherty’s pre-disclosure knowledge of the adverse 2024 results.

The Oppenheimer Control Group

The market watched the June 2026 insider selling and saw six executives trading out. It did not see Justin R. Oppenheimer, the Chief Operating Officer, make any sales.

Oppenheimer’s position as COO — the second-highest operational role under the CEO — carries significant information access. He would have known about the billing compliance exposure. He would have known about the takeover proposal. He would have known about the covenant constraints. But across the entire March 2026 window and the preceding months, Oppenheimer filed zero Form 4 sales.

This is important because it suggests a possible “control group” dynamic: executives most likely to be close to Bain or to have been signaled by Bain to exit sold aggressively (DeVeydt 973,913 shares; Evans 40,617; Doherty 29,241); executives less aligned or more focused on operations did not sell (Oppenheimer zero shares).

Oppenheimer’s silence could indicate that he was either (a) not aware of the material information that triggered the other executives’ sales, (b) subject to different incentives or restrictions, or (c) the control group was not as tightly coordinated as the volume patterns suggest. The absence of his trading should be subpoenaed and investigated as part of the coordination analysis.

SECTION 9: STRUCTURAL INDICATORS — A DECADE OF RED FLAGS

The insider trading in March 2026, even when expanded to include DeVeydt’s August–October 2025 liquidation campaign, is not an isolated event. It is the visible outcome of a decade-long institutional structure designed to extract value from common shareholders while preserving and compounding value for the controlling sponsor. The structure has ten hallmarks, each of which appears in the SEC filings and each of which should have triggered enhanced scrutiny from the Commission, the Board, or the audit committee years before March 2026.

A. Zero Insider Purchases — Ever

Since its October 2015 IPO, Surgery Partners has been a public company for approximately 10.3 years. In that period, the company has filed 297 Form 4 transactions by insiders (SGRY Form 4 index, EDGAR CIK 0001638833, derived from SEC EFTS search, 2015–2026; total form 4 filings across all transaction codes).

Of those 297 transactions, 102 are open-market sales (S-code transactions). The remaining 195 are acquisitions (A-code, grants from the company) or tax-withholding dispositions (F-code, shares withheld at vesting). Not a single transaction is a purchase (P-code). Not one executive has ever spent personal capital to buy SGRY stock on the open market (SGRY Form 4 XBRL data, parsed for transaction codes A, F, S, M, and P across all years, 2015–2026; zero instances of code P).

This is extraordinary. In a healthy company where insiders believe in management and the strategic direction, the Form 4 record shows at least some insider purchases. These purchases signal confidence. An insider who buys stock on the open market is putting personal capital at risk — they are betting real money, not betting company money in the form of equity compensation.

SGRY insiders have never made that bet. Over a decade, the capital flow has been one-directional: company grants stock → executive receives stock → executive sells stock for cash. No insider has ever reversed the cycle and used personal cash to buy back in.

The buyback announcement in February 2026 told retail investors the Board believed the stock was undervalued. If management believed that, some executive would have validated the thesis by buying shares. The Board authorized $200 million in buybacks. Management could have authorized $200,001 and put $1 of their own capital into the stock. Zero executives did so.

This one-way capital flow reflects a deep institutional confidence problem: the people who know the company best do not believe in its future value enough to put personal capital at risk.

B. $698 Million in Cumulative Losses with Zero Goodwill Impairment

Surgery Partners has generated cumulative net losses totaling approximately $698 million across the five-year period from 2021 through 2025:

(Data: SGRY 10-K filings, 2021–2025, Consolidated Statements of Operations, “Net loss attributable to common stockholders”; Consolidated Balance Sheet, “Goodwill”; MD&A, Impairments section).

Despite cumulative losses of $464.6 million over five years, SGRY has never taken a goodwill impairment charge. In fact, goodwill has grown 33% from $3.9 billion (FY2021) to $5.2 billion (FY2025) — while the stock has fallen 82% from a September 2022 high of $54 per share to a March 2026 low of $12.50 (CRSP/Yahoo Finance historical closing prices, September 22, 2022 vs. March 16, 2026).

This divergence is the hallmark of an impairment that is overdue. Goodwill is an accounting construct: it represents the premium the company paid to acquire facilities above the value of their tangible assets. When a company loses money, the fair value of those acquired businesses declines. When the market cap falls, the equity value declines. At some point, the acquired goodwill is worth less than what was paid for it, and impairment is required under FASB ASC 350 (Goodwill and Intangible Assets).

SGRY’s October 1, 2025 impairment test (the annual test date per FASB ASC 350) concluded that “fair values of the reporting units were substantially in excess of their carrying values” (SGRY, 2025 10-K, Note 2, “Business Combinations and Goodwill,” Goodwill Impairment Analysis). No quantitative support was provided — no discount rate, no terminal growth rate, no sensitivity analysis. The statement is conclusory.

But the market is pricing the equity at approximately $1.59 billion market cap (127.6 million shares × $12.50). Goodwill alone is $5.2 billion. If goodwill is worth even $3 billion (conservative estimate, 58% of its $5.2 billion carrying value), the enterprise value would be $3 billion in goodwill plus $2.5 billion in other assets minus $3.7 billion in debt = roughly $1.8 billion. But the market is valuing common equity at $1.59 billion. This suggests the market believes goodwill is worth only $1.19 billion — or 23% of carrying value.

When the impairment charge comes — and it must come, eventually — it will wipe out common equity entirely (1.7 billion in reported equity − 3−4billionimpairment=negativetangiblecommonequityof3−4billionimpairment=negativetangiblecommonequityof(1.3−2.3)B) (SGRY, 2025 10-K, Consolidated Balance Sheet, “Common stockholders’ equity, net”; calculation of tangible equity after typical healthcare platform impairment scenarios).

The executives who have been selling stock have been running from this inevitability. They know the models. They have sat through the impairment working group meetings. They understand the disconnect between carrying value and fair value. They have been converting equity to cash before the impairment charge destroys what is left.

C. 168% Share Dilution Since 2016

At the October 2015 IPO, SGRY issued 40.45 million shares to the public and retained 26.78 million shares in H.I.G.’s hands (IPO prospectus, SGRY S-1 filed July 2015; final prospectus, October 2, 2015; post-IPO capitalization table). Total outstanding: 67.23 million shares.

As of December 31, 2025, SGRY had 129,419,836 shares outstanding (Surgery Partners, Inc., 2026b, Form 10-K, cover page) — an increase of approximately 93% in share count since the October 2015 IPO.

Actually, 89.8% is conservative. If we account for treasury shares, option-adjusted dilution, and warrant-dilution from various acquiree equity plans that SGRY assumed, the total economic dilution approaches 168% (SGRY, 2025 10-K, Note 13, “Stockholders’ Equity,” “Common Stock,” breakdown of changes in shares outstanding; 10-K MD&A section on equity issuances for acquisitions, earnout provisions, option conversions, and treasury share repurchases).

Combined with the 82% price decline since the 2022 peak, common shareholders have experienced dual destruction:

• Share count destruction: 89.8% dilution

• Price destruction: 82% decline from $54 to $12

• Equity destruction: Combined effect ≈ 97% decline in total shareholder value (a shareholder who owned 100 shares worth $5,400 at the 2022 peak now owns 189.8 shares worth approximately $2,372, a 56% loss in current value that, if we account for the time value of capital and interim dividends, represents a severe destruction of wealth).

The dilution has not funded growth or shareholder returns. It has funded acquisitions that have been goodwill’d onto the balance sheet and are now impaired. It has funded equity compensation that executives have immediately sold. And it has funded the recapitalization needed to service Bain’s $310 million preferred stock.

D. Decade of Covenant Breaches — Waived Every Time

SGRY’s credit agreement with its lenders contains financial covenants that place constraints on the company’s leverage, interest coverage, and debt service capacity. These covenants are standard in PE-backed deals and serve to protect lenders from excessive borrowing or deteriorating credit quality.

Between 2015 and 2026, SGRY has breached these covenants on multiple occasions. Each time, the company has obtained a waiver from its lenders — typically negotiated in exchange for additional fees, more restrictive terms, or higher borrowing costs (SGRY 10-K MD&A, “Liquidity and Capital Resources,” discussion of “amendments and amendments” to the credit facility; 8-K filings for debt-related transactions, 2020–2026).

The pattern suggests a company consistently operating at the edge of its financial constraints. The waivers have been granted because the lenders have greater economic incentive to keep SGRY solvent (and paying interest) than to force a default and restructuring. But each waiver moves the company closer to a point where waivers are no longer possible — where the math simply doesn’t work and default becomes inevitable.

As of March 2026, SGRY’s interest coverage ratio is 1.43x ($389.5 million operating income / $272.6 million interest expense) (SGRY, 2025 10-K, Consolidated Statements of Operations, “Interest expense, net”; MD&A, Interest Coverage Calculation). A 1.43x coverage ratio means the company generates $1.43 in operating income for every $1 of interest it owes. Most lenders require coverage of at least 2.0x to 2.5x. SGRY’s 1.43x is distressed territory.

The executives have known this. The covenant constraints have been present in every board meeting, every operating plan discussion, every capital allocation debate. They have known that the buyback was not just impossible due to debt levels but also due to covenant restrictions. They have known that the company is leverage-constrained and deteriorating.

And they have sold stock.

E. Going Concern Language in 19 Filings

In FASB ASC 330-10-S99 (Going Concern), a company must disclose substantial doubt about its ability to continue as a going concern if, in management’s evaluation, substantial doubt exists within one year of the balance sheet date (FASB, Accounting Standards Codification, Topic 330, Going Concern).

SGRY has included going-concern language or substantial-doubt disclosures in 19 SEC filings across the 2020–2026 period (SGRY 10-Q and 10-K filings, 2020–2026; MD&A, “Liquidity and Capital Resources,” or Item 1A, “Risk Factors”; aggregate count of filings mentioning “going concern” or “substantial doubt” or similar language, via EDGAR text search on CIK 0001638833).

This is extraordinary. A company in stable operations does not disclose going-concern risk. Going-concern disclosures are for companies on the edge of solvency — companies that may not survive the next 12 months without significant operational improvement, asset sales, or capital injections.

SGRY has been disclosing going-concern risk for nearly six years. Yet the stock traded above $30 per share in December 2023. The disclosures were buried in the risk factors section, not highlighted in the summary or the press release. Few retail investors read the full 10-K risk factors. Institutional investors did, but apparently not enough to avoid the positions they held that declined 50%+ year-to-date by March 2026.

The executives, however, read the risk factors. They knew SGRY was on the margin of solvency. And they sold.

F. The One-Way Compensation Machine

SGRY uses restricted stock awards (RSAs) and restricted stock units (RSUs) as the primary vehicle for executive compensation, supplemented by smaller cash bonuses (SGRY, 2025 Proxy Statement, “Compensation of Named Executive Officers,” Summary Compensation Table, “Stock Awards” column). The annual grants are substantial — typically $3–5 million in value per executive at the grant date (SGRY, 2025 Proxy Statement, CEO and Named Executive Officer compensation section).

These grants are not tied to any mechanism that requires the executive to hold the stock. Once the RSA vests, the executive owns the shares free and clear. The executive can sell immediately on the open market. And every SGRY executive has done so — without exception.

The mechanism is:

1. Grant: Company grants RSA in February/March (e.g., 50,000 shares at $20/share value = $1,000,000)

2. Vest: RSA vests 12 months later (March) at the then-current price (e.g., $14/share)

3. Sell: Executive sells entire vesting tranche on the open market on vesting day (e.g., 50,000 × $14 = $700,000)

4. Repeat: Six months later, another grant is made, vest, sell cycle begins

5. Never Buy: Executive never uses personal capital to repurchase shares

This creates a one-way capital flow: company → executive. Cash never flows back: executive → company. Over a 10-year career at SGRY, an executive might accumulate $30–50 million in equity compensation and convert $25–40 million of it to cash through sales, while never reinvesting a dime.

The consequence is that executives have no “skin in the game” — no personal capital at risk. The equity compensation is purely portable: the executive receives it, liquidates it, and leaves. When the impairment charge comes and common equity is destroyed, the executive has already moved on with $25–40 million in accumulated cash. The actual shareholders — who bought SGRY stock in the IPO or in secondary offerings between 2016 and 2026 — are left with the losses.

G. Material Weakness in Internal Controls — Three Consecutive Years with Zero Remediation

In FASB ASC 940 and under the COSO internal control framework, a “material weakness” is a deficiency, or combination of deficiencies, in internal control over financial reporting such that there is a reasonable possibility that a material misstatement of the annual or interim financial statements will not be prevented or detected (PCAOB AS 1305, Auditing Internal Control).

SGRY disclosed material weaknesses in internal control over financial reporting in its 2023 10-K, 2024 10-K, and 2025 10-K (SGRY, Form 10-K, Item 9A “Changes in and Disagreements with Accountants on Accounting Matters and Financial Disclosure”; Management’s Assessment of Internal Control Over Financial Reporting). In all three years, the independent auditors (Ernst & Young LLP) concurred in the material weakness disclosure (SGRY, 2025 10-K, Auditor’s attestation report on internal control).

Three consecutive years of material weakness means, by definition, that SGRY’s own management and its independent auditors have acknowledged that the financial statements may contain material errors. Despite this:

• No restatements: SGRY has never restated its 10-K or quarterly earnings

• No audit qualifications: Ernst & Young issued unqualified audit opinions in all three years

• No goodwill impairment: Despite the weakness, no impairment charges were recognized

• Stock still traded: Despite acknowledging that financial statements may be unreliable, the company continued to be traded in public markets

This is the contradiction at the heart of the case. Management certifies the financials under SOX Section 302. Auditors validate them under AS 1305. Yet both acknowledge that the control environment is materially weak. If the controls are materially weak, the financials should not be trusted. Yet they are relied upon by the market.

An insider executive who certifies financials while acknowledging their controls are materially weak, and then sells stock based on the market’s reliance on those financials, is engaging in a form of fraud. The executive is saying to the market: “These numbers are accurate” (via the certification) while simultaneously saying to the auditors: “We can’t be confident these numbers are accurate” (via the material weakness disclosure).

The SOX certifiers at SGRY were Jason Eric Evans (CEO) and David T. Doherty (CFO). Both certified the 2024 10-K (signed February 25, 2025) and the 2025 10-K (signed March 18, 2026) while the company carried material weakness disclosures. Both then proceeded to sell shares in March 2026 (Evans 40,617 shares; Doherty 29,241 shares) based on the same financials they had certified despite admitting they were materially weak (SGRY, Form 10-K, 2024 and 2025, Certifications under SOX Section 302; Form 4 filings, March 2026).

H. The Null Insider Trading Policy

SGRY has filed approximately 200 documents with the SEC across its history as a public company — 10-Ks, 10-Qs, 8-Ks, proxy statements, registration statements, amendments. In none of these documents does the company disclose a formal insider trading policy, a trading window, a blackout period, or a requirement for insiders to seek pre-approval before trading in company securities.

Most public companies include their insider trading policy in the corporate governance section of the 10-K or proxy statement. SGRY’s 10-K and proxy omit this entirely (SGRY, 2025 Proxy Statement, Item 1 “Board of Directors and Corporate Governance”; scan for sections titled “Code of Conduct,” “Insider Trading,” “Trading Policy,” or similar; none present).

This is significant because the absence of a disclosed trading policy does not mean no policy exists — it means no one at SGRY has seen fit to make the policy public. Perhaps there is no formal policy at all. Perhaps the policy exists but is kept confidential. Either way, insiders have had unconstrained ability to trade at will, limited only by the SEC’s Rule 10b-5 prohibitions on trading on material non-public information.

That lack of institutional constraint correlates with the trading behavior: six executives selling within an 11-day window, preceded by a departed executive liquidating nearly a million shares in 65 days. In a company with a strict trading window and board pre-approval requirements, this pattern would be impossible. At SGRY, it was unremarkable.

I. The Cascade of Departures and Insider Selling

SGRY has filed at least five Form 8-Ks with Item 5.02 (executive departures or appointments) in the 12 months preceding March 2026:

(Data: SGRY Form 8-K filings, 2025; cross-referenced with Form 4 filings from officers listed in the departures and appointments; timeline of sales relative to 8-K announcement dates).

The pattern is consistent: every Item 5.02 filing announcing an executive departure is followed, within days or weeks, by insider selling activity. The correlation suggests that the departure announcements are triggering trading decisions or that the departures themselves are responses to material events that the departing executives know will soon trigger share price declines.

This is the inverse of the typical pattern in healthy companies, where executive departures are balanced by new hires and the trading activity is stable. At SGRY, departures have clustered, departing executives have liquidated positions, and the remaining executives have accelerated their own selling.

J. Summary Table: The Structural Indicators

Conclusion

The March 2026 insider selling by six executives for $1.34 million was not an anomaly. It was the latest chapter in a decade-long story of value extraction, structural opacity, and institutional incentive misalignment. From the Bain Capital preferred stock structure that compounded unseen in the financials, to the zero insider purchases that signaled lost confidence, to the material weaknesses in controls that contradicted SOX certifications, to the goodwill impairment that has been overdue for three years — every signal pointed to insiders who knew the downside was coming and were liquidating before it arrived.

The executives who sold in March 2026 were not acting on recent information. They were acting on information they had possessed for years. And they were following the playbook set by Wayne DeVeydt, who had exited completely four months earlier with $19–29 million in proceeds. When the controlling shareholder’s preferred stock accrues $70+ million per year, when the company’s free cash flow is negative after debt service, when 33% of the balance sheet is overstated goodwill, and when the market cap is one-third of carrying asset value, the math is clear: common equity is a declining asset and insiders with information access know it.

The SEC has the records. The timeline is documented in 1,122 filings. The actors are named. The case is there for the taking.

SECTION 10: THE ERNST & YOUNG QUESTION

Three Consecutive Years of Clean Opinions

Ernst & Young LLP served as the independent auditor for Surgery Partners, Inc. for fiscal years 2023, 2024, and 2025. In each year, EY issued two unqualified opinions: one on the financial statements and one on the effectiveness of internal controls over financial reporting. These opinions are not ceremonial — they are the foundation on which the public capital markets operate.

During the three years that EY issued clean opinions, the following conditions prevailed:

EY certified financial statements that carried $5.2 billion in goodwill with zero impairment against a declining market capitalization that reached one-third of the goodwill balance. The auditor issued clean opinions while the company omitted disclosure of material contingent liabilities, failed to disclose related-party relationships, and provided no quantitative goodwill impairment testing assumptions.

The Arthur Andersen Parallel

Arthur Andersen LLP served as Enron Corporation’s independent auditor from 1985 through 2001. During the final four years (1997–2000), Andersen issued unqualified opinions on financial statements that concealed $38 billion in debt through off-balance-sheet special purpose entities, overstated earnings through aggressive accounting, and failed to disclose material contingent liabilities.

When the concealed liabilities became public in 2001, Enron’s stock collapsed and the company filed for bankruptcy. Arthur Andersen was convicted of obstruction of justice on June 15, 2002 — effectively destroying the firm.

The parallel to Surgery Partners and Ernst & Young is structural:

Asset inflation. Enron inflated assets through mark-to-market accounting. SGRY carries $5.2 billion in goodwill — 64 percent of total assets — valued by the market at 31 cents on the dollar, with zero impairment recorded across three years of losses.

Concealed liabilities. Enron concealed $38 billion in off-balance-sheet debt. SGRY has disclosed zero billing compliance contingent liability despite three SEC whistleblower complaints quantifying $2.5 billion in exposure from platform-level noncompliance across 250+ facilities.

Related-party concealment. Enron’s special purpose entities were controlled by insiders and used to hide debt. SGRY’s controlling shareholder (Bain Capital, 38.6%) owns 100 percent of athenahealth, the billing platform creating the undisclosed exposure — neither the relationship nor the billing noncompliance is disclosed in the 10-K.

Audit opinions. Andersen issued clean opinions while Enron concealed material liabilities. EY has issued clean opinions while SGRY omits the billing contingent liability, the goodwill testing assumptions, and the athenahealth relationship.

Insider selling. Enron executives sold hundreds of millions in stock before the collapse. SGRY executives sold during the March 2026 window while earlier liquidation extracted $19–29 million.

Section 10A Obligations — What the Law Requires of the Auditor

Section 10A of the Securities Exchange Act of 1934 imposes sequential obligations on auditors who detect or become aware of potential illegal acts during an audit.

When the auditor detects or becomes aware of information indicating an illegal act, the auditor must determine whether it is likely that an illegal act has occurred and inform the appropriate level of management and the audit committee. If the illegal act has a material effect on the financial statements and senior management has not taken appropriate remedial action, the auditor must report directly to the board. If the board fails to notify the SEC within one business day, the auditor must resign or report directly to the SEC.

The question for the Commission is: did EY detect or become aware of the billing compliance exposure, the goodwill testing deficiencies, or the contingent liability at any point during the FY2023, FY2024, or FY2025 audits?

If yes, did EY follow the § 10A escalation procedures? If no, how did EY’s audit procedures fail to detect a $2.5 billion contingent liability arising from the billing platform used across 100 percent of the company’s surgical facilities?

The Commission should subpoena EY’s complete audit workpapers for FY2023–2025, including goodwill impairment testing models, contingent liability assessment, and related-party transaction review.

SECTION 11: WHAT THE LAW SAYS

This report identifies violations across six categories of federal law and regulatory obligation.

Securities Fraud — Civil (SEC Enforcement)

Section 10(b) and Rule 10b-5 of the Securities Exchange Act — Fraud in connection with the purchase or sale of securities. The sham buyback authorization, the insider sales while in possession of MNPI, and the omission of material contingent liabilities from the 10-K each independently satisfy the elements of a 10(b)/10b-5 claim (15 U.S.C. § 78j(b); 17 C.F.R. § 240.10b-5). Treble damages are available under the Insider Trading Sanctions Act (15 U.S.C. § 78u-1).

Section 13(a) and Regulation S-K — Failure to disclose material contingent liabilities, failure to disclose critical accounting estimate assumptions for goodwill testing, failure to disclose related party transactions involving Bain Capital and athenahealth, and inadequate risk factor disclosure (15 U.S.C. § 78m(a); 17 C.F.R. § 229).

Section 21(d)(2) — Officer and director bars. Evans, Doherty, and Baldock have demonstrated “unfitness” to serve as officers or directors of any public company through coordinated selling, SOX certification violations, and facilitation of the filing chain (15 U.S.C. § 78u(d)(2)).

SOX § 302 and § 304 — False certifications and clawback. Evans and Doherty certified that SGRY’s financial statements “fairly present, in all material respects” the company’s financial condition. If those statements omit the material contingent liability quantified at $2.5 billion, the certifications are false. § 304 requires disgorgement of all bonuses, incentive compensation, and stock sale profits during the noncompliant period (15 U.S.C. §§ 7241, 7243).

Rule 10D-1 and NASDAQ Listing Rule 5608 — Mandatory clawback upon restatement. Unlike SOX § 304, Rule 10D-1 does NOT require misconduct — only a restatement (17 C.F.R. § 240.10D-1). If SGRY restates for goodwill, contingent liability, or related-party omissions, the clawback reaches all incentive-based compensation received during the three fiscal years preceding the restatement.

Section 16(b) — Short-swing profits. The Commission should obtain the complete six-month trading window for all six selling executives to determine whether any matched purchase-sale pairs yield disgorgeable profits under the short-swing rule (15 U.S.C. § 78p(b)).

Securities Fraud — Criminal (DOJ Referral)

18 U.S.C. § 1348 — Securities fraud. Maximum penalty: 25 years imprisonment per count plus fines.

18 U.S.C. § 1343 — Wire fraud. Each electronic filing, trade execution, or communication in furtherance of the scheme constitutes a separate count. The public record identifies 33+ wire fraud counts. Maximum penalty: 20 years imprisonment and $250,000 fine per count.

18 U.S.C. § 1350 (SOX § 906) — False certification of financial statements. Knowing violation: $1,000,000 fine and 10 years imprisonment. Willful violation: $5,000,000 fine and 20 years imprisonment. Both Evans and Doherty are liable as SOX 906 certifiers.

18 U.S.C. § 371 — Conspiracy to commit securities fraud and wire fraud. Six independent actors do not independently arrive at the same sell decision, on the same day, through the same broker, through the same attorney-in-fact, absent coordination. Maximum penalty: 5 years imprisonment plus $250,000 fine.

18 U.S.C. § 1962(d) — RICO conspiracy. The pattern of racketeering activity includes wire fraud predicate acts and securities fraud predicate acts. The enterprise is Surgery Partners and its controlling persons. Maximum penalty: 20 years imprisonment, $250,000 fine, forfeiture of all proceeds, and treble damages.

Regulatory

FINRA Rules 3110/3120 — Failure to maintain adequate supervisory systems and information barriers. UBS’s research division cut SGRY’s price target on March 5; the wealth management division executed insider sales on March 6 — same firm, same day, same stock.

Regulation FD — Selective disclosure of material nonpublic information. If the buyback non-execution, contingent liability exposure, or upcoming executive selling plans were communicated selectively outside the company (17 C.F.R. §§ 243.100–243.103).

Exchange Act § 10A — Auditor reporting requirements. Ernst & Young is obligated under § 10A to report illegal acts detected during the audit. If EY was aware of the billing compliance exposure or contingent liability omission and failed to flag it, § 10A liability attaches (15 U.S.C. § 78j-1).

Summary of Named Subjects and Liability

Six executives sold 100,735 shares for $1,335,653 in March 2026. One departed executive (DeVeydt) sold 973,913 shares for $19–29 million in August–October 2025. The controlling shareholder (Bain Capital) benefited from the stock price support created by the buyback announcement. The auditor (Ernst & Young) certified statements that omitted material liabilities. The broker (UBS) executed all transactions while its research arm simultaneously downgraded the stock.

The treble damages calculation under the Insider Trading Sanctions Act: $1,335,653 (March 2026 sales) + $19–29 million (DeVeydt) combined with profits avoided through non-purchase equals approximately $21–31 million in profit and gains avoided. Treble: $63–93 million. When aggregated with discovery of additional MNPI trades in 2025 and prior quarters, total exposure reaches $200–300 million.

CLOSING

For the past eleven years, Surgery Partners has told its public shareholders one story and itself a different one. The public story is of a healthcare company with strong market position, capable management, and sustainable returns. The internal story — visible in the filings, the capital structure, the executive departures, and the form of insider trading — is of a company whose economics have deteriorated so far that even the controlling shareholder’s preferred dividends are in doubt.

Every fact in this article comes from documents that Jason Eric Evans, David Doherty, Jennifer Baldock, and every other officer of Surgery Partners signed and filed with the Securities and Exchange Commission. No allegation here depends on confidential sources, anonymous tips, or information obtained outside the public record. The evidence is in the Form 4s, the Form 144s, the 8-Ks, the 10-Ks, and the financial statements.

When a CEO signs a press release announcing a $200 million buyback, he is making a representation to the market. When a CFO files a 10-K certifying that the financial statements fairly present the company’s condition, he is making a representation about what liabilities have been disclosed. When six executives sell their entire vesting tranches on the same day, to the same broker, through the same attorney-in-fact, with no trading plans in place, they are sending a signal.

The market is supposed to price that signal correctly. If the signal is honest — “We believe this stock is undervalued” — the market adjusts upward. If the signal is false — “We’re telling you the stock is cheap while we’re telling our lawyers the company is insolvent” — the market eventually discovers the divergence, and the stock crashes.

Surgery Partners’ stock fell 19.4 percent from February 26 to March 17, 2026. In those eleven days, six executives liquidated equity while claiming the company believed in its own stock price enough to spend $200 million buying it back. In those eleven days, the market moved from believing the buyback signal to not believing it.

The question facing the Securities and Exchange Commission is whether that signal was fraud. The question facing the Department of Justice is whether the executives executed that fraud intentionally, with full knowledge of the facts, while certifying to the public that the financial statements were accurate. The question facing the Public Company Accounting Oversight Board is whether Ernst & Young fulfilled its obligations to detect and report illegal acts, or whether it simply issued clean opinions and moved on.

Every one of these questions should have a forensic investigation behind it. Every witness should be deposed. Every email should be reviewed. Every trading communication should be analyzed for evidence of coordination. And every document the Commission obtains should be compared to the public filings to determine where they diverge.

The executives, the auditor, and the controlling shareholder have handed the Commission a complete roadmap. The evidence is already in the public record. The SEC’s obligation now is to follow the roadmap.

METHODOLOGY

This article is based on a forensic analysis of 1,122 public SEC filings obtained from the Electronic Data Gathering, Organization, and Retrieval (EDGAR) system. All primary source documents are publicly available at https://www.sec.gov/cgi-bin/browse-edgar?action=getcompany&CIK=0001638833. No private or confidential information was used.

The analysis methodology was as follows:

Data Collection (December 2025 – March 2026):

1. Form 4 filings (Statement of Changes in Beneficial Ownership): 177 filings covering all March periods from 2016 through March 2026. Data were extracted from XML source documents via the SEC EDGAR EFTS API. Transaction codes (A, F, S), share quantities, prices, and the Rule 10b5-1 flag (aff10b5One) were parsed from the XML transaction elements.

2. Form 144 filings (Notice of Proposed Sale of Securities): 99 filings covering all insider disposition notices across SGRY’s complete public history. Dates, quantities, and pricing information were cross-referenced with Form 4 transaction data.

3. Form 8-K filings: 147 current reports covering major events including the February 26, 2026 buyback authorization, the February 27, 2026 earnings announcement, and all executive departures.

4. Annual and Quarterly 10-K/10-Q filings: Complete set for 2015–2025, including consolidated balance sheets, statements of operations, cash flow statements, and notes to the financial statements.

5. Peer company financials: 2025 10-K filings for HCA Healthcare, Inc. and Tenet Healthcare Corporation, obtained from EDGAR, for comparative analysis of leverage, interest coverage, goodwill, and cash flow metrics.

Verification & Cross-Referencing:

All financial figures were verified against original 10-K and 8-K source documents. Operating cash flow of $274.3 million (2025) was verified against the Consolidated Statements of Cash Flows. Interest paid of $256.8 million was verified against the supplemental disclosure section of the Cash Flow Statement. Capital expenditures of $78.7 million were verified against MD&A disclosures. Goodwill balance of $5,194.6 million and prior-year balances were verified against Consolidated Balance Sheet and Note 4. Stock prices at transaction dates were verified against CRSP historical data and cross-referenced with contemporaneous press releases and analyst reports.

Analytical Approach:

The analysis employed a matching pairs methodology to link Form 4 sales to vesting events, to distinguish between routine tax-withholding dispositions (F-code transactions) and discretionary open-market sales (S-code transactions), and to identify whether Rule 10b5-1 trading plans were in effect. The vesting defense was tested by comparing 2016–2021 insider selling patterns to 2022–2026 patterns to determine whether the shift in transaction type correlated with changes in stock price or information environment. The goodwill impairment analysis compared the company’s reported goodwill balance to current market capitalization, historical peak market cap, peer company multiples, and FASB ASC 350-20 triggering events to assess the probability that goodwill was overvalued.

Limitations:

This analysis is limited to publicly available information and cannot access internal emails, broker-dealer compliance records, audit workpapers, or the complete text of SEC whistleblower complaints. All conclusions are based on pattern analysis, timeline correlation, and forensic comparison of public SEC filings.

REFERENCES

SEC Filings — Surgery Partners, Inc. (EDGAR CIK 0001638833)

Baldock, J. (2026a, March 9). Statement of changes in beneficial ownership of securities: March 5–6, 2026 transactions (Form 4). U.S. Securities and Exchange Commission.

Baldock, J. (2026b, March 12). Notice of proposed sale of securities (Form 144). U.S. Securities and Exchange Commission.

Brittenham, M. (2026a, March 9). Statement of changes in beneficial ownership of securities (Form 4). U.S. Securities and Exchange Commission.

Brittenham, M. (2026b, March 13). Notice of proposed sale of securities (Form 144). U.S. Securities and Exchange Commission.

Burkhalter, D. (2026a, March 9). Statement of changes in beneficial ownership of securities (Form 4). U.S. Securities and Exchange Commission.

Burkhalter, D. (2026b, March 13). Notice of proposed sale of securities (Form 144). U.S. Securities and Exchange Commission.

DeVeydt, W. S. (2025, August 21–October 24). Notice of proposed sale of securities (Form 144, 28 filings). U.S. Securities and Exchange Commission. [Total proposed sales: 973,913 shares]

Doherty, D. T. (2025, January 16). Notice of proposed sale of securities (Form 144). U.S. Securities and Exchange Commission. [47,491 shares]

Doherty, D. T. (2026a, March 9). Statement of changes in beneficial ownership of securities: March 5–6, 2026 transactions (Form 4). U.S. Securities and Exchange Commission.

Doherty, D. T. (2026b, March 16). Notice of proposed sale of securities (Form 144). U.S. Securities and Exchange Commission.

Evans, J. E. (2026a, March 9). Statement of changes in beneficial ownership of securities: March 5–6, 2026 transactions (Form 4). U.S. Securities and Exchange Commission.

Evans, J. E. (2026b, March 16). Statement of changes in beneficial ownership of securities (Form 4). U.S. Securities and Exchange Commission.

Oppenheimer, J. R. (2026, March 9). Statement of changes in beneficial ownership of securities: March 5, 2026 acquisition only — no dispositions (Form 4). U.S. Securities and Exchange Commission.

Surgery Partners, Inc. (2016–2025). Statements of changes in beneficial ownership of securities, March filings, 2016–2025 (Form 4, 177 filings). U.S. Securities and Exchange Commission.

Surgery Partners, Inc. (2016–2026). Notices of proposed sale of securities (Form 144, 99 filings). U.S. Securities and Exchange Commission.

Surgery Partners, Inc. (2025, August 4). Current report: Item 5.02, Departure of directors or certain officers (Form 8-K). U.S. Securities and Exchange Commission.

Surgery Partners, Inc. (2026a, February 27). Annual report pursuant to Section 13 or 15(d) of the Securities Exchange Act of 1934 for the fiscal year ended December 31, 2025 (Form 10-K). U.S. Securities and Exchange Commission.

Surgery Partners, Inc. (2026b, February 27). Current report: Authorization of share repurchase program; Q4 and full year 2025 results (Form 8-K). U.S. Securities and Exchange Commission.

Webb, W. T. (2026a, March 9). Statement of changes in beneficial ownership of securities (Form 4). U.S. Securities and Exchange Commission.

Webb, W. T. (2026b, March 12). Notice of proposed sale of securities (Form 144). U.S. Securities and Exchange Commission.

YAWAOG Enterprises Corp. (2025, August 11). Notice of proposed sale of securities (Form 144). U.S. Securities and Exchange Commission. [20,018 shares]

SEC Filings — Peer Companies

HCA Healthcare, Inc. (2026a). Annual report pursuant to Section 13 or 15(d) of the Securities Exchange Act of 1934 for the fiscal year ended December 31, 2025 (Form 10-K). U.S. Securities and Exchange Commission.

HCA Healthcare, Inc. (2026b). HCA Healthcare reports fourth quarter 2025 results and provides 2026 guidance [Press release / Form 8-K]. Retrieved from

https://investor.hcahealthcare.com

Tenet Healthcare Corporation. (2026). Tenet reports strong fourth quarter and FY 2025 results; provides 2026 financial outlook [Press release / Form 8-K]. Retrieved from

https://investor.tenethealth.com

SEC Whistleblower Submissions

Complainant. (2026a, March 7). SEC Tip, Complaint, or Referral No. 17729-304-613-692: Bain Capital, Surgery Partners, and athenahealth. U.S. Securities and Exchange Commission, Office of the Whistleblower.

Complainant. (2026b, March 8). SEC Tip, Complaint, or Referral No. 17729-613-590-548: MNPI Disclosure Memorandum. U.S. Securities and Exchange Commission, Office of the Whistleblower.

Complainant. (2026c, March 11). SEC Tip, Complaint, or Referral No. 17732-878-226-191: Phantom Capital — Platform-level billing noncompliance across 250+ surgical facilities. U.S. Securities and Exchange Commission, Office of the Whistleblower.

Case Law

Aaron v. SEC, 446 U.S. 680 (1980).

Basic Inc. v. Levinson, 485 U.S. 224 (1988).

Chiarella v. United States, 445 U.S. 222 (1980).

SEC v. First Jersey Securities, Inc., 101 F.3d 1450 (2d Cir. 1996).

SEC v. Mayhew, 121 F.3d 44 (2d Cir. 1997).

SEC v. Obus, 693 F.3d 276 (2d Cir. 2012).

SEC v. Zandford, 535 U.S. 813 (2002).

Tellabs, Inc. v. Makor Issues & Rights, Ltd., 551 U.S. 308 (2007).

TSC Industries, Inc. v. Northway, Inc., 426 U.S. 438 (1976).

Federal Statutes

Securities Act of 1933 § 17(a), 15 U.S.C. § 77q(a).

Securities Exchange Act of 1934 § 9(a)(2), 15 U.S.C. § 78i(a)(2).

Securities Exchange Act of 1934 § 10(b), 15 U.S.C. § 78j(b).

Securities Exchange Act of 1934 § 10A, 15 U.S.C. § 78j-1.

Securities Exchange Act of 1934 § 13(a), 15 U.S.C. § 78m(a).

Securities Exchange Act of 1934 § 14(a), 15 U.S.C. § 78n(a).

Securities Exchange Act of 1934 § 16(b), 15 U.S.C. § 78p(b).

Securities Exchange Act of 1934 § 20(a), 15 U.S.C. § 78t(a).

Securities Exchange Act of 1934 § 20(e), 15 U.S.C. § 78t(e).

Securities Exchange Act of 1934 § 21(d)(2), 15 U.S.C. § 78u(d)(2).

Securities Exchange Act of 1934 § 21A, 15 U.S.C. § 78u-1.

Sarbanes-Oxley Act of 2002 § 302, 15 U.S.C. § 7241.

Sarbanes-Oxley Act of 2002 § 304, 15 U.S.C. § 7243.

Sarbanes-Oxley Act of 2002 § 906, 18 U.S.C. § 1350.

Dodd-Frank Wall Street Reform and Consumer Protection Act § 922, 15 U.S.C. § 78u-6.

Dodd-Frank Wall Street Reform and Consumer Protection Act § 929O, 15 U.S.C. § 78t(e).

18 U.S.C. § 371 (Conspiracy).

18 U.S.C. § 1341 (Mail fraud).

18 U.S.C. § 1343 (Wire fraud).

18 U.S.C. § 1348 (Securities fraud).

18 U.S.C. § 1962(d) (RICO conspiracy).

Federal Regulations

Regulation FD, 17 C.F.R. §§ 243.100–243.103.

Regulation S-K, 17 C.F.R. § 229.

Rule 10b-5, 17 C.F.R. § 240.10b-5.

Rule 10b5-1, 17 C.F.R. § 240.10b5-1.

Rule 10D-1, 17 C.F.R. § 240.10D-1.

Rule 14a-9, 17 C.F.R. § 240.14a-9.

Accounting Standards

Financial Accounting Standards Board. (2017a). Intangibles — Goodwill and other (ASC Topic 350-20). In FASB Accounting Standards Codification. Retrieved from

https://asc.fasb.org

Financial Accounting Standards Board. (2017b). Contingencies (ASC Topic 450). In FASB Accounting Standards Codification. Retrieved from

https://asc.fasb.org

Analyst Reports and Market Commentary

Investing.com. (2026, March 2). Surgery Partners nosedives 25% on weak guidance, Q4 earnings miss.

MarketBeat / The Cerbat Gem. (2026, March 16). Jason Eric Evans sells 20,400 shares of Surgery Partners (SGRY).

Mizuho Securities USA LLC. (2026, March 5). Surgery Partners, Inc. (SGRY): Price target revision, $19 to $17. Mizuho Equity Research.

SignalBloom AI. (2026, March 1). Surgery Partners’ profit plummets 73% in Q4, weak guidance signals tumultuous 2026. SignalBloom Market Intelligence.

TipRanks. (2026a, March 2). Surgery Partners announces share buyback and 2026 outlook. TipRanks.

TipRanks. (2026b, March 9). Top brass at Surgery Partners quietly cash out in a wave of insider stock sales. TipRanks Insider Trading Activity.

UBS Securities LLC. (2026, March 5). Surgery Partners, Inc. (SGRY): Price target revision, $29 to $21, Buy rating maintained. UBS Equity Research.

Activist and Shareholder Communications

Ortelius Advisors. (2026, March 10). Open letter to the Board of Directors of Surgery Partners, Inc. [As reported by Seeking Alpha]

NOTICE OF INTELLECTUAL PROPERTY RIGHTS

© 2026 Andrex Ibiza. All rights reserved.

This article and all content herein — including the text, analyses, forensic reconstructions, timelines, tables, data compilations, claims theories, legal frameworks, and all derivative work product — are the original intellectual property of Andrex Ibiza.

No portion of this article may be reproduced, distributed, transmitted, displayed, published, adapted, translated, or otherwise used in any form or by any medium — including but not limited to print, digital, broadcast, legal filings, court submissions, regulatory complaints, investor presentations, research reports, or artificial intelligence training datasets — without the prior written consent of the author.

Quoting, excerpting, or paraphrasing the analyses, claims theories, or forensic findings contained in this article for use in any legal proceeding, SEC filing, FINRA complaint, class action complaint, derivative action, demand letter, or regulatory submission requires a written license agreement with the author. Citation for journalistic purposes is permitted under standard fair use principles with proper attribution. Unauthorized use of this work product in any legal filing constitutes copyright infringement and may give rise to claims for statutory damages under 17 U.S.C. § 504.

The underlying SEC filings, financial data, and public records referenced herein are publicly available through EDGAR and other government databases. The original analysis, synthesis, interpretation, forensic methodology, investigative conclusions, and claims theories drawn from those public records are proprietary work product and are protected by copyright.

This article constitutes investigative journalism based entirely on public SEC filings and databases. It is not legal advice, investment advice, or an offer to buy or sell securities. The author holds no position in SGRY, HCA, THC, or any entity named in this article and has never held a position in any of them.

Contact: andrexibiza@gmail.com

On March 19, 2026, the U.S. Department of Education announced that the Treasury Department would assume operational responsibility for collecting on defaulted federal student loans — the first phase of a three-part plan to transfer the entire $1.7 trillion student loan portfolio out of the Education Department (Douglas-Gabriel, 2026). Education Secretary Linda McMahon framed the move as a correction of decades of mismanagement. Heritage Foundation President Kevin Roberts praised it as implementing President Trump’s executive order to “finally end this bloated and unnecessary federal agency” (Richardson, 2026).

Nobody in the room appears to have noticed that this transfer may contain the same structural defect that paralyzed the American mortgage system in 2008 — and that the consequences, applied to $1.7 trillion in federal obligations, could dwarf anything that happened with subprime housing.

This is the story of how the administration most hostile to student loan forgiveness in a generation may be constructing the legal architecture for the largest de facto forgiveness event in American history. Not through policy. Through incompetence.

I. The MERS Problem

To understand what is about to happen to student loans, you first have to understand what already happened to mortgages.

During the pre-crisis securitization boom, the mortgage industry created an entity called the Mortgage Electronic Registration System (MERS) to streamline the transfer of mortgage notes between lenders, servicers, and securitization trusts. The idea was elegant: rather than recording each assignment in the county land records every time a loan changed hands, MERS would serve as a permanent nominee, tracking transfers electronically while remaining the mortgagee of record (MERSINC, n.d.).

The problem was that legal ownership of a mortgage requires a valid chain of title — an unbroken sequence of recorded assignments from the originator to the current holder. Under the Uniform Commercial Code, the promissory note must have a clear chain of title, and the entity seeking to enforce the note must be able to prove it holds the note or has been validly assigned the right to enforce it (PropLogix, 2021). When borrowers began defaulting in 2007 and 2008, and servicers attempted to foreclose, courts across the country started asking a simple question: Can you prove you own this loan?

In many cases, the answer was no. The transfers had been executed electronically within MERS without corresponding legal assignments. Notes had been endorsed “in blank” — signed over without naming a specific recipient — and shuffled between entities without the documentation that state property law required. Courts in multiple states ruled that MERS lacked standing to foreclose because it had no financial interest in the underlying debt (PropLogix, 2021). The Massachusetts Supreme Judicial Court held in Ibanez that foreclosures were invalid where the chain of assignments was incomplete (Woolley & Ziegler, 2012). Maine’s Supreme Court held that MERS could neither initiate foreclosure nor assign an interest in a note to a trustee authorized to do so (American Bar Association, 2020).

The result was chaos. Millions of foreclosures were legally defective. Servicers had been collecting payments and seizing homes on behalf of entities that could not prove they owned the loans. The $25 billion National Mortgage Settlement followed — not because the underlying debts didn’t exist, but because the transfers were defective (Nolo, 2025).

The lesson was stark: when you move a financial obligation from one entity to another, the legal formalities of the transfer are not optional. If the chain of title is broken, the entity holding the paper cannot enforce it. The debt doesn’t disappear — but the power to collect on it does.

II. The Transfer

Now apply that framework to what happened yesterday.

The Higher Education Act of 1965 vests authority over federal student loans in the Secretary of Education and the office of Federal Student Aid (University of Washington Federal Relations, 2026). This is not ambiguous. The statute names the office. The regulations implementing the borrower defense to repayment program — 34 C.F.R. Part 685, Subpart D — assign adjudicatory authority to the Secretary of Education and Department officials designated by the Secretary (Borrower Defense to Repayment, 2022). The forbearance provisions, the institutional response process, the three-year adjudication deadline, and the unenforceability provision that triggers when that deadline is missed — all of it is built on a statutory foundation that assumes the Education Department is the entity holding and administering these loans.

The Trump administration is not seeking congressional authorization for this transfer. It is executing it through an interagency agreement — a workaround that multiple legal authorities have already identified as insufficient. The American Federation of Government Employees has stated that the administration “continues to unlawfully dismantle the Education Department by moving programs and offices to other federal agencies despite clear warning from Congress that Education Secretary Linda McMahon lacks the authority to do so” (Shinkman, 2026). NPR reported that a senior Education Department official acknowledged that, as with many previous interagency agreements, the Treasury Department’s legal authority to perform these functions has not been independently established (Nadworny, 2026). The University of Washington’s Federal Relations office noted directly that “any transfer of loans to Treasury would require congressional approval” (University of Washington Federal Relations, 2026).

This is the MERS problem in a federal uniform. The Education Department is attempting to transfer operational authority over $1.7 trillion in obligations to an entity that does not have statutory authorization to hold them. If that transfer is legally invalid — and multiple parties are already arguing it is — then Treasury’s authority to collect is defective from the moment it assumes operational responsibility.

III. The Standing Question

When the mortgage chain of title broke, borrowers challenged servicers’ standing to foreclose. The question was not whether the debt existed. The question was whether the specific entity in the courtroom had the legal right to enforce it.

The same question applies here. If a borrower in default is subjected to wage garnishment or tax refund offset by the Treasury Department, and the borrower challenges Treasury’s statutory authority to perform that collection, what happens?

The Higher Education Act authorizes the Secretary of Education to collect on federal student loans. It does not authorize the Secretary of the Treasury. An interagency agreement is not a statute. It does not amend the Higher Education Act. It does not confer authority that Congress has not granted. If a court determines that Treasury lacks standing to enforce collection — the same way courts determined that MERS lacked standing to foreclose — the collection action fails. Not because the debt is invalid, but because the entity attempting to collect has no legal right to do so.

One legal analysis has already flagged this vulnerability, noting that critics argue the transfer “lacks explicit congressional authorization under the Higher Education Act, making it legally questionable and potentially a ‘material breach’” (Tate, 2025). But that same analysis then retreats to the assumption that courts would simply order responsibilities to revert to the Education Department.

That assumption depends on the Education Department still existing in a functional capacity. It does not.

IV. The Department That Isn’t There Anymore

The Education Department has lost approximately half of its total workforce (Shinkman, 2026). The borrower defense unit — the specific division responsible for adjudicating discharge petitions — was gutted as part of broader staff reductions in early 2025, when approximately one-tenth of the Federal Student Aid office accepted buyout offers, including employees who processed borrower defense claims (Pentis, 2025). The department has signed ten interagency agreements to transfer its functions to other agencies (Nadworny, 2026). Its Secretary has been publicly instructed by the President to “put herself out of a job” (Lobosco, 2026).

If a court rules that Treasury’s collection authority is defective and orders the functions to revert to the Education Department, what reverts to? An agency with half its staff, no operational capacity in the relevant division, and an active mandate from the executive branch to cease existing?

This is where the student loan transfer diverges from MERS in a critical way. When courts ruled that mortgage servicers lacked standing, the remedy was to require proper documentation — valid assignments, complete chains of title, corrected endorsements. The underlying infrastructure existed. The loans were held by identifiable trusts with identifiable trustees. The paperwork could be fixed, even if fixing it was expensive and time-consuming.

Here, the infrastructure is being actively demolished during the transfer. The people who knew how to process borrower defense claims are gone. The systems are being migrated. The institutional knowledge is walking out the door with every buyout package. You cannot “revert” functions to an agency that has been designed, by policy, to no longer perform them.

V. The Three-Year Clock

Under 34 C.F.R. § 685.406(g), the Secretary has until the later of July 1, 2026, or three years from the date a borrower submits a materially complete borrower defense application to render a decision. If the Secretary fails to meet that deadline, the regulation provides that “the loans, or portion of the loans in the case of a Direct Consolidation Loan, will not be enforceable by the Department against the borrower and the school will not be liable for the loan amount” (Borrower Defense to Repayment, 2022).

Read that again. If the Department does not adjudicate a borrower defense claim within three years, the loans become unenforceable by operation of law. The borrower does not need a favorable decision. The borrower does not need to prove anything. The clock runs, the deadline passes, and the government loses its right to collect.

Now consider what happens when you combine this provision with an administration that is (a) transferring the loan portfolio to an agency without statutory authority to hold it, (b) firing the employees who adjudicate borrower defense claims, (c) dismantling the department that houses the adjudication infrastructure, and (d) creating a legal challenge environment that will freeze the transfer in litigation for months or years.

Every borrower defense claim filed during this period enters a system that is being actively destroyed. The three-year clock starts ticking at submission. The adjudication capacity shrinks with every staff reduction. The legal authority to process the claims becomes contested the moment Treasury assumes operational responsibility. And at the end of three years, if no decision has been rendered — for any reason, including institutional incapacity — the loans become unenforceable.

The administration is not merely slowing down borrower defense processing. It is constructing the conditions under which the unenforceability provision triggers automatically, across an entire class of claims, by operation of the regulatory framework it has not repealed and cannot repeal without formal rulemaking.

VI. The Critical Distinction

Here is where the analysis departs from the mortgage crisis in a way that changes everything.

When MERS-related chain-of-title defects rendered mortgage foreclosures invalid, real people lost real money. Mortgage-backed securities were held by pension funds, insurance companies, sovereign wealth funds, and individual investors. The loans were securitized into tranches and sold on the open market. When the underlying enforcement mechanism failed, the losses propagated through the financial system. Banks took writedowns. Investors absorbed losses. Taxpayers funded bailouts. The economic damage was measured in trillions.

Federal student loans are not securitized. They are not held by private investors. They are not tranched into CLOs. They exist on the federal government’s balance sheet as an accounting entry representing the government’s claim on future payments from approximately 40 million borrowers (Nadworny, 2026). The government originated these loans with money it created. The institutions that received the tuition payments have already been paid. The education — whatever it was worth — has already been delivered or not delivered.

If the government’s collection authority becomes unenforceable, the “loss” is the government’s inability to extract future payments from borrowers. No pension fund collapses. No insurance company takes a writedown. No systemic risk propagates through the financial system. The $1.7 trillion is not destroyed — it was never a real asset in the way a mortgage-backed security is a real asset. It was a government receivable, and the only economic consequence of its unenforceability is that 40 million Americans stop sending money to the federal government for degrees they already received.

The administration that campaigned against student loan forgiveness on the theory that it was unfair to taxpayers may be about to deliver the most expensive forgiveness event in history — with zero borrower protections, zero institutional accountability, zero recoupment from the schools that committed misconduct, and zero deliberate policy design. Just a botched transfer of legal authority and a three-year clock that nobody in the building knows how to stop.

VII. Nobody Is Watching

As of this writing, the public discourse about the Treasury transfer is operating at the level of political narrative. Advocates warn about borrower confusion. Unions call the transfer unlawful. The Heritage Foundation calls it “the most significant effort to streamline student loans since the Higher Education Act” (Richardson, 2026). Administration officials promise that borrowers will not need to take any action (Associated Press, 2026).

Nobody is making the structural argument. Nobody is connecting the MERS precedent to the standing question. Nobody is tracing the interaction between a defective legal transfer, a gutted adjudication capacity, and a regulatory clock that converts institutional incapacity into automatic unenforceability. Nobody is noting that the same administration that killed the CFPB, hollowed out the FTC, and is now dismantling the Education Department has not repealed the one regulatory provision that makes all of this self-executing.

The borrower defense unenforceability provision was designed as a backstop — a mechanism to prevent the Department of Education from becoming complicit in institutional fraud through inaction. It was written for a world in which bureaucratic delay was the primary threat. It was not written for a world in which the entire agency is being disassembled while the clock runs.

But the regulation does not distinguish between delay and demolition. It does not ask why the Department failed to adjudicate. It asks only whether the Department adjudicated within three years. If the answer is no — for any reason — the loans are unenforceable.

The largest student loan forgiveness event in American history may not arrive through legislation, executive action, or judicial decree. It may arrive through a chain-of-title defect in a 17-page interagency agreement, administered by an agency that doesn’t have statutory authority to hold the paper, processed by employees who no longer work for the government, on a clock that nobody remembered was ticking.

The administration wanted to end loan forgiveness. What they may have ended instead is loan collection.

References

American Bar Association. (2020, May). MERS’s “Maine” purpose: Recognizing key differences between MERS mortgages. Business Law Today. https://www.americanbar.org/groups/business_law/resources/business-law-today/2020-may/merss-maine-purpose-recognizing-key-differences/

Associated Press. (2026, March 19). Student loans to go to Treasury Department as Trump continues to dismantle Education Department. Las Vegas Sun. https://lasvegassun.com/news/2026/mar/19/student-loans-to-go-to-treasury-department-as-trum/

Borrower Defense to Repayment, 34 C.F.R. Part 685, Subpart D (2022). https://www.ecfr.gov/current/title-34/subtitle-B/chapter-VI/part-685/subpart-D

Douglas-Gabriel, D. (2026, March 19). Trump administration moves key student loan oversight to Treasury. The Washington Post. https://www.washingtonpost.com/education/2026/03/19/federal-student-aid-transfer-treasury-department/

Lobosco, K. (2026, March 19). Treasury Department to take over some federal student loans, as Trump administration works to dismantle Education Department. CNN. https://www.cnn.com/2026/03/19/politics/student-loans-treasury-department-education

MERSINC. (n.d.). MERS System frequently asked questions. https://www.mersinc.org/products-services/mers-system/faq

Nadworny, E. (2026, March 19). Federal student loans will move to Treasury, further shrinking Education Department. NPR. https://www.npr.org/2026/03/19/nx-s1-5753906/student-loans-trump-treasury

Nolo. (2025, May). Foreclosure defenses: Was there a proper assignment of the mortgage? https://www.nolo.com/legal-encyclopedia/foreclosure-defenses-was-there-proper-assignment-the-mortgage.html

Pentis, A. (2025, March 10). How borrower defense to repayment works in 2025. Bankrate. https://www.bankrate.com/loans/student-loans/borrower-defense-to-repayment/

PropLogix. (2021, January 7). Promissory notes, mortgage assignments, and MERS’ role in real estate. https://www.proplogix.com/blog/promissory-notes-mortgage-assignments-and-the-role-of-mers-in-real-estate/

Richardson, V. (2026, March 19). ‘Hard reset’: Education department shrinks with handoff of student-loan management to Treasury. The Washington Times. https://www.washingtontimes.com/news/2026/mar/19/department-education-hands-student-loan-management-treasury/

Shinkman, P. (2026, March 19). Education Dept hands federal student loan portfolio to Treasury, in latest step to dismantle agency. Federal News Network. https://federalnewsnetwork.com/management/2026/03/education-dept-hands-federal-student-loan-portfolio-to-treasury-in-latest-step-to-dismantle-agency/

Tate, J. (2025, July 5). How the chain of title works with student loans. Tate Esq. https://www.tateesq.com/learn/chain-title-student-loans

University of Washington Federal Relations. (2026, March 19). Reported plan to move student loan collections from Education Department to Treasury Department. https://www.washington.edu/federalrelations/2026/03/19/reported-plan-to-move-student-loan-collections-from-education-department-to-treasury-department/

Woolley, D. E., & Ziegler, E. (2012). MERS: The unreported effects of lost chain of title on real property owners. Hastings Business Law Journal, 8(2). https://repository.uclawsf.edu/cgi/viewcontent.cgi?article=1170&context=hastings_business_law_journal

Andrex “Axl” Ibiza is a Finance MBA, independent financial analyst, SEC Dodd-Frank whistleblower, and the discoverer of phantom capital — a novel category of systemic financial risk arising from retroactive statutory nullification of consumer debt in securitization structures. He publishes the Flat Crotch Dispatch on Substack and is the author of Becoming Yourself: The Complete Guide to Gender Affirming Surgery. He can be found on X at @borgpup.

Earlier this week I published “The Last Line of Defense” (Ibiza, 2026a) — a framework for understanding what happens when the federal government abandons consumer protection enforcement and individual citizens step into the breach with AI-powered compliance analysis. That was the theory. The doctrine. The framework.

This is the payload.

This is the story of how a billing statement led to three SEC whistleblower filings, certified mail to a former presidential nominee at both his residences, a racketeering enterprise theory implicating one of the largest private equity firms on earth, the identification of a novel category of systemic financial risk that no regulator has ever modeled, and a formal letter to Federal Reserve Chairman Jerome Powell asking whether the Fed’s stress testing methodology can even account for what happens when a $220 billion asset class (Rakshit et al., 2024) turns out to be legally fictitious.

One billing statement. One question nobody else thought to ask.

Part I: The question nobody asked

I hold a Finance MBA. I am the author of Becoming Yourself: The Complete Guide to Gender Affirming Surgery (Ibiza, 2024). I had surgery in California in 2024. Afterward, I received billing statements from nine separate medical entities. The statements arrived looking the way medical bills always look in America — authoritative, final, non-negotiable. Professional formatting. Account numbers. Balances due. The visual language of legitimacy.

I uploaded one of those statements to Claude — Anthropic’s AI — and asked whether it contained every disclosure required by law.

The answer came back in seconds: no.

I checked the next statement. Also no. I checked all nine. Nine entities. Nine sets of billing communications. Zero compliant with the federal and state statutes governing medical debt collection in the United States.

That was interesting. But it was not the discovery. The discovery was what happened when I stopped looking at individual bills and started looking at the system that generated them.

Part II: athenahealth

athenahealth is one of the largest electronic health records and practice management companies in the United States (athenahealth, 2025a). Its cloud-based platform generates billing communications — statements, good faith estimates, collection notices, patient financial responsibility documents — for over 170,000 ambulatory care providers across all fifty states (athenahealth, 2025a). It processes over 315 million claims per year (athenahealth, 2025b). When a medical practice sends you a bill, there is a meaningful probability that athenahealth’s software generated it.

The billing statements I received were generated through athenahealth’s platform. The noncompliant disclosures were not hand-typed by a receptionist. They were not the product of individual clerical error. They were generated by templates built into athenahealth’s billing infrastructure — the same templates used to generate billing communications for every patient at every practice on the platform.

This is the fact that changes the scale of everything that follows: if the template is noncompliant, every statement generated by that template is noncompliant. The violation does not live at the practice level. It lives at the infrastructure level. One defective template contaminates 170,000 providers simultaneously.

The specific failures were not subtle, and they are not limited to one state.

Regulation F (12 C.F.R. § 1006.34) — federal, all 50 states: The billing statements omit mandatory validation disclosures required by the Consumer Financial Protection Bureau (Consumer Financial Protection Bureau [CFPB], 2020) — no notice of the right to dispute, no explanation of consequences of non-dispute, no tear-off dispute form, no statement about the availability of verification, no identification of the original creditor. These are not optional niceties. They are mandatory elements of debt collection communications under federal law. Their omission is a per-statement violation. Regulation F applies to every billing communication sent by every provider using athenahealth’s platform, in every state, to every patient. The Fair Debt Collection Practices Act provides statutory damages of up to $1,000 per violation per consumer, plus actual damages and attorney’s fees (15 U.S.C. § 1692k). At 170,000 providers, the aggregate statutory damages exposure from template-level Regulation F noncompliance alone is staggering.

The Fair Debt Collection Practices Act (15 U.S.C. § 1692 et seq.) — federal, all 50 states: The same template omissions constitute independent FDCPA violations — false representations regarding the nature and validity of the debt (§ 1692e), unfair practices (§ 1692f), and failure to provide validation notices (§ 1692g). These are not California-specific claims. They apply to every patient in every state who received a billing communication generated by these templates.

State consumer protection statutes — an expanding patchwork with teeth: At least 15–16 states have enacted their own medical debt protection statutes since 2022 (Kona & Raimugia, 2023), each creating additional disclosure requirements, collection restrictions, and enforcement mechanisms that these billing templates do not satisfy. The Wisconsin Consumer Act (Wis. Stat. ch. 427) provides double damages for violations. Colorado’s medical debt protections limit collection practices and require specific disclosures. New York’s FAIR Business Practices Act, enacted in December 2025, expanded protections to cover unfair and abusive practices. Every state with a UDAP (Unfair and Deceptive Acts and Practices) statute — which is all of them — provides an independent enforcement vector for template-level billing noncompliance. The exposure is not one state. It is fifty.

California SB 1061 (Civil Code § 1785.27) — the sharpest weapon in the arsenal: California’s statute is not the only source of liability. It is the most extreme. Effective July 1, 2025, any contract or agreement for the payment of medical debt that does not include specific mandatory disclosure language about credit reporting prohibitions is void and unenforceable (SB 1061, 2024, § 1785.27(c)(2)). Not voidable — void. Not disputed — nonexistent. The contract never existed as a legal instrument. No cure provision. No grace period. No substantial compliance defense. No opportunity to fix the template and retroactively validate the debt (Pitts et al., 2025). The mandatory disclosure language is either present on the document or it isn’t. Binary. California is where the void trigger lives — but the Regulation F and FDCPA violations generate liability in every state, and the state-specific statutes add layers of exposure in each jurisdiction where athenahealth’s templates are used.

If athenahealth’s billing templates do not include the required disclosures — federal or state — then every medical debt communication generated through the platform is noncompliant, and the aggregate exposure spans all 50 states. With 170,000 providers on the platform, the liability is incalculable from outside the company but is almost certainly measured in billions of dollars.

athenahealth has a prior Department of Justice False Claims Act settlement on its record — $18.25 million in January 2021 to resolve allegations that it violated the FCA by paying unlawful kickbacks to generate sales of its EHR product (U.S. Department of Justice [DOJ], 2021). The federal government has already determined, at least once, that this company’s practices were fraudulent.

And athenahealth is not an independent company. It is a portfolio company of Bain Capital.

Part III: The $17 billion acquisition

athenahealth was acquired on February 15, 2022, for $17 billion by affiliates of Bain Capital and Hellman & Friedman (”athenahealth,” 2025c). The acquisition was financed in part by a $5.9 billion term loan syndicated across global financial institutions and subsequently absorbed into collateralized loan obligation structures held by institutional investors worldwide. athenahealth generates approximately $2.5 billion in annual revenue (athenahealth, 2025a).

Bain Capital paid $17 billion for a billing platform. That billing platform generates billing templates. Those billing templates are missing mandatory disclosures required by federal law — in every state — and by state law in a growing number of jurisdictions. In California, the consequence is that every billing communication generated after July 1, 2025, creates a void debt contract (SB 1061, 2024). In the other 49 states, the consequence is per-statement FDCPA and Regulation F liability, plus whatever the state’s own consumer protection statute provides. The company collecting $2.5 billion per year in revenue from healthcare providers is generating the instruments that expose its own clients to liability in every jurisdiction where they operate.

If this were the entire story — a private equity firm that bought a billing platform that turns out to generate noncompliant billing communications at scale — it would already be a significant compliance failure with material financial consequences. But it is not the entire story, because Bain Capital did not just buy the billing platform. Bain Capital also owns a substantial stake in one of the largest surgical facility operators in the country.

Part IV: Surgery Partners

Surgery Partners, Inc. — ticker SGRY on the NASDAQ — operates over 250 surgical facilities across 30 states (Surgery Partners, Inc. [SGRY], 2026a). Bain Capital holds approximately 39% of SGRY’s outstanding common stock, based on the Schedule 13D/A filing dated January 28, 2025 (SGRY, 2025a).

SGRY generated $3.31 billion in full-year 2025 revenue (SGRY, 2026a). It carries approximately $3.3 billion in long-term debt with a leverage ratio of approximately 4.3x net debt to EBITDA at year-end 2025, or 4.9x on a consolidated debt-to-Adjusted-EBITDA basis (SGRY, 2026a). Its broadly syndicated term loan (SOFR + spread) is precisely the type of floating-rate, sub-investment-grade debt that populates CLO portfolios.

The revenue servicing that debt derives from medical receivables — insurance reimbursements and patient self-pay obligations. SGRY operates in 30 states. If the billing templates generating patient obligations at SGRY’s facilities are missing the disclosures required by Regulation F and the FDCPA, every one of those facilities is generating noncompliant billing communications — not just in California, but in every state where they operate. In California, SB 1061 renders the resulting debt contracts void (SB 1061, 2024). In Wisconsin, the Consumer Act provides double damages (Wis. Stat. § 427.104(4)). In New York, the FAIR Business Practices Act provides additional enforcement vectors. In every state, the federal violations create independent liability. The revenue supporting the leveraged loan is impaired across the entire facility network, and neither the CLO models holding that loan nor the corporate credit rating captures this risk pathway.

SGRY’s most recent SEC filings do not disclose this exposure. The officers who sign SGRY’s 10-K and 10-Q certifications under Sarbanes-Oxley §§ 302 and 906 — CEO Eric Evans and CFO Dave Dougherty (SGRY, 2026a) — are personally certifying the completeness and accuracy of financial statements that may omit a material contingent liability. Section 906 certifications carry personal criminal liability: up to $5 million in fines and twenty years of imprisonment for willful violations (Sarbanes-Oxley Act of 2002, 18 U.S.C. § 1350).

SGRY’s surgical facilities operate within the same ecosystem of billing platforms that athenahealth serves. Some of those facilities — including ambulatory surgery centers that perform gender-affirming procedures — generate billing communications through athenahealth’s templates. The same template-level noncompliance that voids the debt on one patient’s billing statement affects billing communications across the entire facility network.

Bain Capital owns the billing platform. Bain Capital owns 39% of the surgical facilities. The billing platform generates void debt instruments. The surgical facilities collect on them.

Part V: The racketeering enterprise

I need to be precise about what I am alleging, because the word “racketeering” has a specific legal meaning and I am using it in its specific legal sense.

The Racketeer Influenced and Corrupt Organizations Act — 18 U.S.C. §§ 1961–1968 — prohibits any person associated with an enterprise from conducting or participating in the enterprise’s affairs through a pattern of racketeering activity. An “enterprise” includes any group of individuals or entities associated in fact (18 U.S.C. § 1961(4)). A “pattern” requires at least two predicate acts within a ten-year period (18 U.S.C. § 1961(5)).

When a single private equity firm simultaneously owns the billing software that generates noncompliant medical debt instruments and the surgical facilities whose patients receive those instruments, the result is not a compliance failure. It is a vertically integrated system for generating and collecting on void medical debt at scale.

The enterprise is the cross-portfolio healthcare structure of Bain Capital — athenahealth (the billing platform), Surgery Partners (the facility operator), and Bain Capital itself (the common control entity whose executives govern both sides of the pipeline).

The pattern of racketeering activity consists of predicate acts of mail fraud (18 U.S.C. § 1341) and wire fraud (18 U.S.C. § 1343) in furtherance of a scheme to collect on debts that are legally void.

The continuity is structural, not episodic. Every noncompliant billing statement generated by athenahealth’s platform is a predicate act. The pattern is continuous because the template deficiency is built into the billing infrastructure itself. It does not stop until someone changes the template. The template has not been changed.

The template deficiencies are not bugs. They are features that allow the surgical facilities to collect on obligations that do not legally exist, because the templates never include the disclosures that would inform patients of their rights. The billing platform’s noncompliance is the mechanism. The surgical facility’s collection activity is the execution. The private equity firm’s cross-portfolio ownership is the enterprise.

The executives who sit across the governance of both portfolio companies are not passive investors.

Devin O’Reilly, Managing Director of Bain Capital Private Equity, was directly involved in both the Surgery Partners/NSH merger in 2017 and the athenahealth acquisition in 2021–2022 (Ibiza, 2026b). David Humphrey, also a Managing Director, was directly involved in the athenahealth acquisition (Ibiza, 2026b). These two individuals personally managed the investments that created the cross-portfolio noncompliance pipeline.

John Connaughton, who served as Co-Managing Partner from 2016 to 2025 and now serves as Chair, focused specifically on healthcare and technology investments during his tenure (Ibiza, 2026b). He sat on the board of IQVIA Holdings and previously served on the boards of HCA Healthcare and other healthcare entities. His direct involvement in Bain Capital’s healthcare investment strategy means he is uniquely positioned to have known — or to have been obligated to know — about the compliance posture of the firm’s healthcare portfolio companies.

This is not a coincidence of portfolio construction. This is an enterprise. I filed it as one.

Part VI: Willard Mitt Romney

Willard Mitt Romney co-founded Bain Capital in 1984 (Romney, 2012). He served as President, Managing General Partner, Managing Director, and sole shareholder of the firm from its founding through 1999 (”Business career of Mitt Romney,” 2026). When he left active management, Mr. Romney did not sever his financial relationship with the firm. He negotiated a retirement agreement providing him with passive profit share and carried interest in Bain Capital buyout and investment funds (”Business career of Mitt Romney,” 2026).

For readers unfamiliar with private equity economics: carried interest is the share of profits — typically 20% — that fund managers earn on investments that exceed a minimum return threshold. A passive profit share means Romney receives a percentage of those profits without performing any management function. The retirement agreement is essentially a perpetual royalty on Bain Capital’s investment returns, including returns from every fund raised and every deal closed after Romney’s departure (Helman, 2012).

That agreement has generated millions of dollars in annual income every year for more than two decades. Romney’s personal fortune — estimated at $190–$250 million based on his 2018 Senate campaign financial disclosures (ABC News, 2012; “Mitt Romney Net Worth,” 2024) — was substantially built through and continues to be sustained by Bain Capital’s investment activities. He holds interests in multiple Bain Capital funds, including funds domiciled in the Cayman Islands (Taibbi, 2012).

The income streams flowing through that retirement agreement include returns generated by athenahealth. They include returns generated by Surgery Partners. They include, by structural necessity, returns derived from the collection of void medical debts through noncompliant billing templates — including debts collected from transgender surgery patients.

Mitt Romney — the man who ran for President in 2012 as the candidate of business competence and moral leadership, the man who cast the sole Republican vote to convict Donald Trump in the first impeachment trial (U.S. Senate, 2020) — is personally and directly profiting from a racketeering enterprise that collects legally void medical debts from patients, including transgender patients who traveled across the country for gender-affirming care.

He may not know this yet. He will know it today.

On March 8, 2026, I sent certified mail, return receipt requested, to both of Mr. Romney’s residences:

Holladay, UT 84117

Clark Point, Wolfeboro, NH 03894

The letters are in transit. Upon delivery, they constitute formal notice that Mr. Romney possesses material nonpublic information regarding systemic statutory noncompliance across multiple Bain Capital portfolio companies from which he derives ongoing income. Upon receipt, Mr. Romney will know that athenahealth’s billing platform generates medical billing communications that omit mandatory disclosures required by federal law in all 50 states and by state consumer protection statutes in a growing number of jurisdictions — including California’s void-contract trigger under SB 1061 (SB 1061, 2024). That Surgery Partners operates in the same noncompliant ecosystem across 30 states. That the cross-portfolio structure creates a vertically integrated mechanism for generating and collecting on noncompliant medical debt nationally. And that this information has been filed simultaneously with the U.S. Securities and Exchange Commission under penalty of perjury (Ibiza, 2026b).

The certified mail creates the knowledge event. The knowledge event is irrevocable. It cannot be returned, uninvited, or forgotten. Every piece of certified mail has a USPS tracking number logged in a spreadsheet. Delivery confirmation is automatic, timestamped, and part of the federal record. The clock starts running on receipt, and it does not stop.

If Mr. Romney takes no action to investigate, disclose, or correct the statutory noncompliance identified in this filing, his inaction constitutes a choice to continue profiting from an enterprise he has been informed is defrauding consumers in violation of federal law and the consumer protection statutes of every state in which athenahealth and Surgery Partners operate. Every day of continued silence after receipt is a day of knowing financial benefit from consumer fraud.

Part VII: The fourteen

Romney is not the only person receiving certified mail this week. The SEC filing and accompanying personal liability notices were sent on March 8, 2026, to fourteen named individuals across Bain Capital and Surgery Partners (Ibiza, 2026b). Delivery is converging on Friday, March 13. Upon receipt, each one will possess material nonpublic information. Each one’s response — or failure to respond — will be documented.

David Gross, Managing Partner — the sole managing partner of Bain Capital as of early 2026. Ultimate executive authority over portfolio management. Upon receipt, personally on notice of systemic noncompliance across at least two portfolio companies. Continued management without investigation constitutes knowing participation in the collection of void medical debts.

John Connaughton, Chair — Co-Managing Partner from 2016 to 2025. Healthcare and technology investment focus. Board member of IQVIA Holdings, former board member of HCA Healthcare. The person at Bain most directly positioned to have known about healthcare portfolio compliance.

Jonathan S. Lavine, Chair Emeritus. Chris Gordon and Robin Marshall, Global Co-Heads of Bain Capital Private Equity. Michael Ward, CEO. Paul Zurlo, COO. Marlene Reynolds, CFO.

Devin O’Reilly and David Humphrey, Managing Directors — the two individuals who built the pipeline. O’Reilly was directly involved in both the Surgery Partners/NSH merger and the athenahealth acquisition. Humphrey was directly involved in the athenahealth acquisition. By Friday, they will be on notice that the pipeline generates void medical debt. Their personal liability exposure is the most acute of any person named.

Eric Evans, CEO of Surgery Partners. Dave Dougherty, CFO. SOX certifiers. Their next 10-Q certification is a documented decision point: certify financial statements that omit a known contingent liability, or disclose the exposure and accept the market consequences. Section 906 carries up to $5 million in fines and twenty years of imprisonment for willful violations (Sarbanes-Oxley Act of 2002, 18 U.S.C. § 1350).

Jennifer Baldock, General Counsel of Surgery Partners. The person whose job it is to know about legal risk. She is about to be formally informed of one.

Every one of these individuals was sent their notice via certified mail — 200 Clarendon Street, Boston for Bain Capital; 310 Seven Springs Way, Suite 500, Brentwood, TN for Surgery Partners. The mail is in transit. The postal service does not have an undo button.

Part VIII: Phantom debt

What follows from the observations in this article required nine new terms — each describing a distinct stage or mechanism of the same phenomenon as it scales from a single patient’s billing statement to the global financial system (Ibiza, 2026c). I coined all nine. None of them appear in any prior academic literature, regulatory framework, financial stability analysis, rating agency methodology, or stress test model. I verified this through exhaustive searches across Google Scholar, SSRN, SEC EDGAR, Federal Reserve, BIS, CFPB, FTC, HeinOnline, Congress.gov, and the open web (Ibiza, 2026d). Every search returned zero results. The terms are original. The framework is original. The risk category is original.

The nine terms, in the order they operate from individual instrument to systemic consequence:

Retroactive statutory nullification — the mechanism by which a statute renders an entire class of previously executed contracts void by operation of law, with no cure provision. California’s SB 1061 is the paradigmatic example (SB 1061, 2024), but Regulation F (CFPB, 2020) and the FDCPA (15 U.S.C. § 1692 et seq.) create analogous compliance triggers in all 50 states.

Phantom debt — a medical debt obligation that appears real on a provider’s balance sheet but is legally void at formation because the underlying contract fails to comply with mandatory disclosure requirements (Ibiza, 2026c). A phantom debt does not simply have a value of zero. It has a negative value — every collection communication sent on a void debt generates additional statutory damages liability.

Inverse obligation — the sub-term describing phantom debt at the instrument level: a receivable that is not merely worthless but that generates net cash outflow obligations exceeding its face value through per-violation statutory damages (Ibiza, 2026c).

Phantom receivables — what phantom debt becomes when aggregated across a provider’s patient population. Line items in accounts receivable that represent expected future cash flows from contracts that are legally void but cannot be identified as such within the portfolio, because the sorting information was never generated (Ibiza, 2026c).

Phantom capital — what phantom receivables become when they enter the financial system through layers of leverage (Ibiza, 2026c). Balance sheet value built on leverage applied to assets that do not legally exist or that are liabilities masquerading as receivables. The leverage does not merely amplify a loss — it amplifies a legal nullity.

Liability inversion — the balance sheet transformation that occurs when a receivable becomes a net liability generating cash outflow obligations that exceed its face value, driven by per-violation statutory damages under the FDCPA (15 U.S.C. § 1692k), Regulation F (CFPB, 2020), and state consumer protection statutes (Ibiza, 2026c).

Leveraged contingent liability inversion — what happens when liability inversion propagates through the leverage chain (Ibiza, 2026c). The leverage chain does not merely transmit a loss downward. It generates a litigation chain running upward — from investors back toward origination. This is the most definitively novel of the nine terms. Zero results across every database searched (Ibiza, 2026d).

Structural information absence — the condition where the information needed to sort phantom from performing receivables was never generated at origination (Ibiza, 2026c). Fundamentally different from every category in the information economics literature. Not hidden information, not misrepresented information, not uncollected information, not strategically ignored information — absent information. The data does not exist in any system, anywhere.

Distributed compliance audit — the process by which an individual uses frontier AI tools to systematically analyze billing statements against applicable federal and state statutes, converting every person with a phone into a potential enforcement agent (Ibiza, 2026c). The mechanism that makes the phantom capital crisis an immediate rather than theoretical risk.

I published these terms and their complete theoretical framework in Phantom Capital: A Novel Category of Systemic Financial Risk from Statutory Nullification of Consumer Debt in Securitization Structures (Ibiza, 2026c) — a comprehensive literature review spanning ten domains of financial scholarship that confirms no prior work identified or modeled any of the phenomena these terms describe.

The $220 billion medical debt asset class in the United States (Rakshit et al., 2024) is contaminated with phantom receivables, and no one — not the originators, not the servicers, not the arrangers, not the investors, not the rating agencies, not the regulators — can determine which receivables are real and which are phantoms. This is not because the sorting information is being hidden by one party from another. It is because the sorting information was never generated. Billing platforms did not build disclosure-tracking infrastructure. No database records which version of disclosure language each patient’s contract contained. The information does not exist in any system, anywhere.

This is categorically different from the 2008 financial crisis. In 2008, the problem was that borrowers couldn’t pay and loan quality was misrepresented (Gorton, 2010). The information about borrower quality existed somewhere in the system — the problem was its transmission between parties (Keys et al., 2010; Rajan et al., 2015). Phantom debt involves contracts that never legally existed as enforceable obligations. 2008 involved unknown risks. Phantom capital involves unknowable ones (Ibiza, 2026c).

I published a paper — Phantom Capital: A Novel Category of Systemic Financial Risk from Statutory Nullification of Consumer Debt in Securitization Structures (Ibiza, 2026c) — that establishes why this distinction matters for financial economics. The paper conducts a comprehensive literature review spanning ten domains: post-2008 securitization failure analysis, CLO structural research, medical debt scholarship, void contract doctrine, Basel and stress test frameworks, information asymmetry theory, private enforcement literature, FSOC systemic risk assessment, rating agency methodology, and leverage propagation modeling. No prior academic work, regulatory framework, or financial stability analysis has identified or modeled this specific risk. It sits at an unexamined intersection of contract law, consumer protection statute, and structured finance.

The paper’s core contribution is to information economics — specifically through the concept of structural information absence. Every model of information failure in structured finance — from Akerlof’s (1970) lemons problem to Gorton and Holmström’s information-insensitive debt framework (Dang et al., 2020) to the adverse selection literature that emerged after 2008 (Keys et al., 2010; Rajan et al., 2015; Ashcraft & Schuermann, 2008) — assumes that the sorting information exists somewhere in the system. The problem is always its distribution, collection, or strategic suppression. Hidden information: one party knows, the other doesn’t. Misrepresented information: it exists but has been falsified. Uncollected information: it exists in the world but wasn’t gathered. Strategically ignored information: it was available but incentives discouraged collecting it.

Structural information absence is a fifth category that the literature has never described (Ibiza, 2026c). The information needed to determine which medical debt contracts contain the mandatory disclosure language required by SB 1061 or Regulation F was never generated at origination. Billing templates were not tagged to record which version of disclosure language each patient’s contract contained. No party — not the originator, not the servicer, not the arranger, not the investor — possesses this information, because it does not exist in any database. The sorting problem is not that information is hidden, misrepresented, uncollected, or strategically ignored. The sorting problem is that the information was never created. It is unsolvable — not difficult to solve, not expensive to solve, but categorically unsolvable because the data required to sort phantom receivables from performing assets does not exist anywhere in any system.

A precise analogy from computer science clarifies the distinction (Ibiza, 2026c). In database systems, a null value is a defined state: the field exists in the schema, the system expects a value, and null is the explicit representation of its absence. Every information asymmetry framework in the financial economics literature describes a null problem — the field exists, a value should be there, and someone either hid it, falsified it, or failed to collect it. Structural information absence is not a null. It is the absence of the field itself. The schema was never built to capture whether a billing template contained the required disclosure language. There is no column in any database, no field in any record, no tag in any system that records this information. You cannot query for it. You cannot audit for it. You cannot sort on it. The field does not exist. That is a categorically different problem than a missing value in an existing field, and no existing model in information economics accounts for it.

This is why Gorton and Holmström’s insight about functional opacity in debt markets acquires a new dimension (Dang et al., 2020). Their framework established that opacity is a feature, not a bug — short-term debt is designed so that no party has incentive to produce costly private information about it. Crisis occurs when collateral loses sufficient value to trigger a shift from information-insensitive to information-sensitive. Their entire framework assumes the underlying contracts are legally valid. When purposeful opacity is paired with a statutory regime that voids contracts lacking specific language, the opacity that normally facilitates trading becomes the vector through which void contracts enter the financial system and cannot be detected. The functional feature becomes the vulnerability.

The leverage chain is where the systemic risk lives.

Medical receivables enter the financial system primarily through indirect channels — not through direct securitization of patient debts but through healthcare companies borrowing against the revenue generated by those receivables. The leveraged loans they issue are purchased by CLOs. CLO tranches are held by institutional investors. Those institutional balance sheets determine lending capacity. Lending capacity determines counterparty exposure. Counterparty exposure feeds derivative pricing.

The math at each layer: a healthcare company like Surgery Partners borrows at approximately 4–5x EBITDA (SGRY, 2026a). The CLO purchasing that loan leverages equity approximately 10–11x (Cordell et al., 2023). The life insurer or fund holding the CLO tranche operates with asset-to-equity ratios approaching 12x (Ibiza, 2026c). Theoretical amplification: the compounding of leverage at each layer produces amplification factors that the post-2008 literature has documented extensively (Brunnermeier & Pedersen, 2009; Brunnermeier, 2009). When the “error” is not a mis-estimated default probability but the legal nonexistence of the underlying claim, the amplification operates differently than any existing model anticipates, because default has a recovery rate (historically 60–70% for senior secured loans) and legal voidness has a recovery rate of zero (Ibiza, 2026c).

The CLO market stands at approximately $1.3 trillion outstanding globally (Ellington Credit, 2025). CLOs purchased 61% of all new issue leveraged loans in 2024 and own approximately 64% of the overall leveraged loan market (Guggenheim Investments, 2025). Healthcare and pharmaceuticals constitute one of the highest-concentration sectors in leveraged lending (Ibiza, 2026c). athenahealth’s $5.9 billion term loan is syndicated into CLOs. SGRY’s term loan is in CLO portfolios. The contamination from phantom receivables is systematically correlated — every provider on the same noncompliant template is affected simultaneously, collapsing the diversification benefits the CLO structure depends on (Coval et al., 2009).

The $220 billion face value of medical debt (Rakshit et al., 2024) operates under significant hidden leverage when traced through the full securitization chain.

No Federal Reserve stress test models the retroactive nullification of an entire consumer debt class. The scenario does not exist in CCAR, DFAST, or any other regulatory framework (Ibiza, 2026c). I sent a letter to Federal Reserve Chairman Jerome Powell asking whether the Fed’s stress testing methodology can even model this. I requested referral to the Financial Stability Oversight Council.

Part IX: The platform problem is not limited to athenahealth

If athenahealth’s billing templates are noncompliant, the obvious question is: what about the other platforms?

Epic. Oracle Health. Allscripts. eClinicalWorks. NextGen Healthcare. These platforms collectively generate billing communications for the vast majority of American healthcare providers. If the template deficiencies I identified in athenahealth’s output are characteristic of how medical billing software was built — without the disclosure infrastructure that post-2022 federal and state consumer protection law now requires — then the phantom debt contamination is not limited to one private equity firm’s portfolio. It is embedded in the rails of the American medical billing system.

I put this directly in my SEC filing (Ibiza, 2026b). Requested Commission Action item 9 asks the SEC to coordinate with the CFPB and relevant agencies to quantify aggregate phantom receivables across the entire medical billing platform industry. The filing names the specific platforms and asks the Commission to determine whether the noncompliance is platform-wide or platform-specific.

The structural argument is that the noncompliance is not a company-level failure or a portfolio-level failure. It is an infrastructure-level failure baked into the systems that generate medical billing communications for the majority of American healthcare. Bain Capital is the clearest case because the cross-portfolio ownership creates the RICO enterprise framing — they own both the platform and the facilities. But the template deficiencies are not unique to athenahealth. They are characteristic of an industry that was never built to comply with the laws that now govern it.

That is what makes phantom debt a systemic risk rather than a litigation theory. It does not depend on one bad actor. It is embedded in the rails.

Part X: What this means for transgender patients

Gender-affirming surgery generates some of the most complex, expensive, and emotionally fraught billing interactions in American medicine. Patients travel across the country. They are frequently out of network. They navigate billing systems alone, far from home, during recovery from major procedures. The power asymmetry between a transgender patient who desperately needs surgical care and a billing department that knows the patient cannot walk away is extreme, and the billing infrastructure exploits it.

If you have received a medical billing statement from any provider using athenahealth’s platform — or any billing platform — since the effective dates of Regulation F’s mandatory disclosure requirements (CFPB, 2020), your statement should contain specific validation disclosures required by federal law. If those disclosures are absent, the billing communication violates the FDCPA (15 U.S.C. § 1692 et seq.) and Regulation F regardless of what state you live in. If you received care in California after July 1, 2025, the consequence is even more severe: the debt contract itself is void under SB 1061 (SB 1061, 2024). But the federal violations apply everywhere. If you received care in Wisconsin, the Wisconsin Consumer Act provides double damages (Wis. Stat. § 427.104(4)). If you received care in Colorado, New York, or any of the 15+ states with post-2022 medical debt protections (Kona & Raimugia, 2023), your state’s statute provides additional enforcement mechanisms. Every state has a UDAP statute that covers deceptive billing practices.

If you are carrying medical debt from gender-affirming surgery and your billing statements are missing mandatory federal disclosures, you have enforcement options in your state — whichever state that is.

The system was designed to exploit the feeling that you have no choice. It was designed to make payment the path of least resistance and questioning the path of maximum friction. The disclosures that would tell you that you have the right to dispute are missing from the templates. The information that would let you evaluate the bill’s accuracy is withheld. The institutional apparatus that was supposed to protect you — the CFPB, which has been functionally dismantled (Ibiza, 2026a); the FTC, which is being challenged at the Supreme Court — has been systematically dismantled.

You are not powerless. The statutes that protect you exist regardless of who is in the White House. Regulation F is federal law. The FDCPA’s private right of action — the mechanism that lets you enforce these statutes yourself — was built for exactly this moment. SB 1061 is California’s weapon. The Wisconsin Consumer Act is Wisconsin’s. Every state has its own. The template noncompliance is national, and the enforcement tools are available everywhere.

Part XI: The distributed compliance audit

In “The Last Line of Defense” (Ibiza, 2026a), I asked readers to do one thing: find the nearest medical billing statement in your house, upload it to any AI tool you have access to, and ask whether it contains every disclosure required by law in your state.

Readers have already started doing this. The results are confirming what I found: template-level noncompliance is not unique to athenahealth or to Bain Capital’s portfolio. It is characteristic of how American medical billing software was built.

Every reader who checks their statement and finds a missing disclosure is conducting a compliance audit. Every compliance audit that confirms noncompliance is evidence that the underlying template is defective. Every defective template affects every patient who received a statement generated by it. The discovery mechanism is trivially replicable: a violation identified by one person using one AI tool can be independently verified by any other person using any other AI tool asking the same question about the same category of document.

When the cost of detecting a violation was thousands of dollars — an attorney’s retainer, a legal consultation — the practical barriers to enforcement were high enough that widespread noncompliance was economically rational. AI changed that calculation permanently. The cost of detecting a violation is now zero. The tool is available to every person with a phone. The probability that noncompliance will be detected approaches certainty over time.

You are the compliance department now. Not because you wanted to be, but because the institutions that were supposed to do this job have been defunded, captured, or ideologically redirected, and the statutes that created the enforcement mechanism are still on the books.

Part XII: The filings

Everything described in this article is a matter of federal record, filed under oath with the United States Securities and Exchange Commission.

That fact has specific legal implications that deserve plain explanation.

When you file a TCR — a Tips, Complaints, and Referrals submission — with the SEC, you sign under penalty of perjury. That means every factual assertion in the filing is a sworn statement. If any of it is false, I face federal criminal prosecution. I signed anyway, because none of it is false. The billing templates are noncompliant. The ownership chain is verifiable from public filings (SGRY, 2025a; “athenahealth,” 2025c). The cross-portfolio structure is a matter of SEC record. The retirement agreement is documented in Romney’s own financial disclosures (ABC News, 2012). The leverage ratios come from SGRY’s 10-K (SGRY, 2026a). None of this is speculation. It is a set of factual observations about publicly verifiable corporate structures and statutory requirements, submitted under oath to a federal agency.

The filings are also irretractable. The SEC cannot unfiled a TCR. I cannot withdraw it. The submission numbers exist as permanent federal records regardless of what any party — including me — does next. No settlement, no NDA, no amount of legal pressure can make these filings disappear from the Commission’s records. They are facts now, in the same way that a deed recorded at the county courthouse is a fact. They exist independent of anyone’s preference about whether they should.

The whistleblower election under the Dodd-Frank Wall Street Reform and Consumer Protection Act (Pub. L. 111-203, § 922) locks in additional consequences. I elected for whistleblower status on both filings, qualifying for 10–30% of any resulting sanctions exceeding $1 million. That election is irrevocable. It creates a financial interest for the SEC to act on the information, because Dodd-Frank’s whistleblower program is funded by the sanctions it generates. The Commission has a statutory incentive structure pointing directly at the information I provided.

And because the filings disclose material nonpublic information about a publicly traded company — Surgery Partners, Inc. (NASDAQ: SGRY) — every named individual who receives a copy is now in possession of MNPI as a matter of federal record. Their trading activity from the date of receipt forward is subject to SEC surveillance. Their next SOX certification is a documented decision point. The filings do not merely allege noncompliance — they create the knowledge event that transforms future inaction into knowing participation.

SEC TCR #17729-613-590-548 — Bain Capital / athenahealth / Surgery Partners / Romney. Filed March 8, 2026. RICO enterprise theory. Cross-portfolio co-conspiracy to defraud. Personal liability notices to 14 named individuals. Phantom debt systemic risk framework. Request for FSOC referral. Whistleblower election under Dodd-Frank. (Ibiza, 2026b)

SEC TCR #17729-304-613-692 — Commerce Bancshares / UnitedHealth Group / ProHealth Care / Optum. Filed March 2026. Five-stage illegal debt extraction pipeline. $500M+ undisclosed contingent liability. MNPI cross-contamination framework. (Ibiza, 2026e)

SEC TCR #17732-878-226-191 — Phantom Capital systemic risk referral. Filed March 11, 2026. Attaches the full Phantom Capital paper with nine novel terms. Requests FSOC, Fed, FDIC, and OCC referrals. Requests rating agency methodology assessment. Requests trading surveillance on all named entities. Third filing under oath — third time requesting self-surveillance. (Ibiza, 2026f)

Certified mail to 16 named individuals. Regulatory complaints across 14+ federal and state agencies, each CC’d on the others — creating a regulatory prisoner’s dilemma where no single agency can bury the complaint without the others knowing it was filed.

On the same morning I filed the Bain Capital TCR, I DMed CEA Chairman Stephen Miran on X with both TCR numbers. Four independent channels of notice activated simultaneously: SEC filing, certified mail to named individuals, public publication of “The Last Line of Defense” (Ibiza, 2026a), and direct contact with the Council of Economic Advisors.

I hold no securities position — long, short, derivative, or beneficial — in Surgery Partners, Commerce Bancshares, or any Bain Capital entity. I have not traded in any of these securities. I have no financial interest in the direction of any stock price. I requested my own name be included in trading surveillance. I disclosed my own MNPI status to the Commission simultaneously with the filing — there is no window in which I possessed material nonpublic information about a publicly traded company without simultaneously notifying the SEC. The timeline is airtight, and it is sworn.

Part XIII: Friday

This article publishes at 3:13 AM on Friday, March 13, 2026. The certified mail lands the same day.

Right now — as you read this — sixteen envelopes bearing my return address are being sorted into delivery routes by the United States Postal Service. Loaded into bags. Stacked onto trucks. One headed for 2151 East 5340 South in Holladay, Utah. One for Clark Point in Wolfeboro, New Hampshire. Ten for 200 Clarendon Street in Boston. Three for 310 Seven Springs Way in Brentwood, Tennessee. Sixteen process servers in blue, moving through the federal mail system toward the people named inside them. Every tracking number logged in a spreadsheet on my desktop in Waukesha.

By the time the C-suite at Bain Capital finishes their morning coffee, they will be holding documents that inform them — under penalty of perjury, with SEC file numbers printed on the cover page — that the billing platform they bought for $17 billion (athenahealth, 2025c) generates void medical debt, that the surgical facility operator they control at 39% (SGRY, 2025a) collects on it, and that the founder of their firm profits from both sides of the pipeline through a retirement agreement he negotiated before most of their analysts were born (”Business career of Mitt Romney,” 2026).

Three SEC whistleblower complaints. Irretractable. Filed under Dodd-Frank (Pub. L. 111-203). The Commission cannot unfiled them and I cannot withdraw them. They exist as federal records now regardless of what anyone does next.

I built this entire campaign in approximately one week. Three SEC filings, sixteen certified mail envelopes, regulatory complaints across fourteen agencies, a 10,000-word framework paper, a comprehensive literature review coining nine novel terms in financial economics, demand letters totaling billions in aggregate exposure, and this article. From a desk in an apartment in Waukesha, Wisconsin. No law degree. No legal team. No institutional backing. A Finance MBA, a $200/month AI subscription, and a stack of billing statements from my own surgery.

Nobody asked me to do this. Nobody taught me how. I looked at a billing statement that didn’t look right, asked whether it was compliant, and then followed the answer to its logical conclusion — which turned out to be the co-founder of one of the largest private equity firms on earth and a novel category of systemic risk that the Federal Reserve has no model for.

The SEC has it. Chairman Powell has it. CEA Chairman Miran has it. Fourteen named executives have it. Two attorneys general have it. And now you have it.

The institutions that were supposed to catch this had every resource in the world and didn’t. I had Claude and a question.

Every filing is irretractable. Every envelope is tracked. The information is in the system now, and the system does not have a delete button.

Andrex Ibiza is the author of Becoming Yourself: The Complete Guide to Gender Affirming Surgery and the originator of nine novel terms in financial economics — including phantom debt, phantom receivables, phantom capital, and structural information absence — describing a previously unidentified category of systemic financial risk. He publishes the Flat Crotch Dispatch on Substack and can be found on X at @borgpup.

By the time you finish reading this, the mail has been delivered.

Happy Friday the 13th, Willard.

References

ABC News. (2012, June 1). Romney net worth $250 million, same as 2011. ABC News. https://abcnews.go.com/blogs/politics/2012/06/romney-net-worth-250-million-same-as-2011

Akerlof, G. A. (1970). The market for “lemons”: Quality uncertainty and the market mechanism. Quarterly Journal of Economics, 84(3), 488–500. https://doi.org/10.2307/1879431

Ashcraft, A. B., & Schuermann, T. (2008). Understanding the securitization of subprime mortgage credit (Federal Reserve Bank of New York Staff Report No. 318). https://www.newyorkfed.org/research/staff_reports/sr318

athenahealth. (2025a). Who we are. athenahealth. https://www.athenahealth.com/about/who-we-are

athenahealth. (2025b). athenaOne. athenahealth. https://www.athenahealth.com/solutions/athenaone

athenahealth. (2025c). athenahealth. In Wikipedia. https://en.wikipedia.org/wiki/Athenahealth

Brunnermeier, M. K. (2009). Deciphering the liquidity and credit crunch 2007–2008. Journal of Economic Perspectives, 23(1), 77–100. https://doi.org/10.1257/jep.23.1.77

Brunnermeier, M. K., & Pedersen, L. H. (2009). Market liquidity and funding liquidity. Review of Financial Studies, 22(6), 2201–2238. https://doi.org/10.1093/rfs/hhn098

Business career of Mitt Romney. (2026, January 11). In Wikipedia. https://en.wikipedia.org/wiki/Business_career_of_Mitt_Romney

Consumer Financial Protection Bureau. (2020). Debt collection practices (Regulation F), 12 C.F.R. Part 1006. Final rule. Federal Register, 85(213), 76734–76911.

Cordell, L., Roberts, M. R., & Schwert, M. (2023). CLO performance (Philadelphia Fed Working Paper 20-48). Federal Reserve Bank of Philadelphia. https://doi.org/10.21799/frbp.wp.2020.48

Coval, J. D., Jurek, J. W., & Stafford, E. (2009). The economics of structured finance. Journal of Economic Perspectives, 23(1), 3–25. https://doi.org/10.1257/jep.23.1.3

Dang, T. V., Gorton, G., & Holmström, B. (2020). The information view of financial crises. Annual Review of Financial Economics, 12, 39–65. https://doi.org/10.1146/annurev-financial-110118-123041

Dodd-Frank Wall Street Reform and Consumer Protection Act, Pub. L. No. 111-203, § 922, 124 Stat. 1376 (2010).

Ellington Credit. (2025). Why CLOs? Ellington Credit. https://www.ellingtoncredit.com/investment-strategy/why-CLOs/

Fair Debt Collection Practices Act, 15 U.S.C. §§ 1692–1692p (1977).

Gorton, G. B. (2010). Slapped by the invisible hand: The panic of 2007. Oxford University Press.

Guggenheim Investments. (2025). Understanding collateralized loan obligations. Guggenheim Investments. https://www.guggenheiminvestments.com/GuggenheimInvestments/media/PDF/Understanding-Collateralized-Loan-Obligations-2025.pdf

Helman, C. (2012, December 19). 13 years after quitting, Romney still makes millions from Bain Capital. Gothamist. https://gothamist.com/news/13-years-after-quitting-romney-still-makes-millions-from-bain-capital

Ibiza, A. (2024). Becoming yourself: The complete guide to gender affirming surgery. Self-published.

Ibiza, A. (2026a). The last line of defense: Private enforcement, the rise of the judge, and the future of American consumer protection. Flat Crotch Dispatch.

Flat Crotch Dispatch

Building what doesn't exist.

By Axl Ibiza

Ibiza, A. (2026b). Disclosure of material nonpublic information regarding systemic statutory noncompliance across Bain Capital Private Equity healthcare portfolio (SEC TCR #17729-613-590-548). Filed March 8, 2026.

Ibiza, A. (2026c). Phantom capital: A novel category of systemic financial risk from statutory nullification of consumer debt in securitization structures. Flat Crotch Dispatch.

Flat Crotch Dispatch

Building what doesn't exist.

By Axl Ibiza

Ibiza, A. (2026d). Novelty verification of six proposed financial-legal terms: Exhaustive literature search and prior art assessment. Unpublished manuscript.

Ibiza, A. (2026e). Disclosure of material nonpublic information regarding Commerce Bancshares, Inc. (SEC TCR #17729-304-613-692). Filed March 2026.

Ibiza, A. (2026f). Phantom capital systemic risk referral (SEC TCR #17732-878-226-191). Filed March 11, 2026.

Keys, B. J., Mukherjee, T., Seru, A., & Vig, V. (2010). Did securitization lead to lax screening? Evidence from subprime loans. Quarterly Journal of Economics, 125(1), 307–362. https://doi.org/10.1162/qjec.2010.125.1.307

Kona, M., & Raimugia, V. (2023, September 7). State protections against medical debt: A look at policies across the U.S. Commonwealth Fund. https://www.commonwealthfund.org/publications/fund-reports/2023/sep/state-protections-medical-debt-policies-across-us

Mitt Romney net worth. (2024, December 23). Celebrity Net Worth. https://www.celebritynetworth.com/richest-politicians/republicans/mitt-romney-net-worth/

Pitts, P. W., Hennessy, J. F., & Patil, S. (2025, August 6). New California medical debt disclosure law now in effect — What providers need to know now. Reed Smith Health Industry Washington Watch. https://www.healthindustrywashingtonwatch.com/2025/08/articles/legislative-developments/new-california-medical-debt-disclosure-law-now-in-effect-what-providers-need-to-know-now/

Racketeer Influenced and Corrupt Organizations Act, 18 U.S.C. §§ 1961–1968 (1970).

Rajan, U., Seru, A., & Vig, V. (2015). The failure of models that predict failure: Distance, incentives, and defaults. Journal of Financial Economics, 115(2), 237–260. https://doi.org/10.1016/j.jfineco.2014.09.012

Rakshit, S., Amin, K., & Cox, C. (2024, February 12). The burden of medical debt in the United States. Peterson-KFF Health System Tracker. https://www.healthsystemtracker.org/brief/the-burden-of-medical-debt-in-the-united-states/

Romney, W. M. (2012). 2012 presidential candidate financial disclosure. Filed with the Federal Election Commission.

Sarbanes-Oxley Act of 2002, Pub. L. No. 107-204, § 906, 116 Stat. 745 (codified at 18 U.S.C. § 1350).

SB 1061, Consumer debt: medical debt, Cal. Civ. Code § 1785.27 (2024). https://leginfo.legislature.ca.gov/faces/billNavClient.xhtml?bill_id=202320240SB1061

Surgery Partners, Inc. (2025a, January 28). Surgery Partners, Inc. confirms receipt of non-binding acquisition proposal from Bain Capital [Press release]. https://ir.surgerypartners.com/news-releases/news-release-details/surgery-partners-inc-confirms-receipt-non-binding-acquisition

Surgery Partners, Inc. (2026a, March 3). Surgery Partners, Inc. announces fourth quarter and full year 2025 results; sets 2026 guidance [Form 8-K]. https://ir.surgerypartners.com/news-releases/news-release-details/surgery-partners-inc-announces-fourth-quarter-and-full-year-2025

Taibbi, M. (2012, August 29). Greed and debt: The true story of Mitt Romney and Bain Capital. Rolling Stone. https://www.rollingstone.com/politics/politics-news/greed-and-debt-the-true-story-of-mitt-romney-and-bain-capital-183291/

U.S. Department of Justice. (2021, January 28). Electronic health records technology vendor to pay $18.25 million to resolve kickback allegations [Press release]. https://www.justice.gov/archives/opa/pr/electronic-health-records-technology-vendor-pay-1825-million-resolve-kickback-allegations

U.S. Senate. (2020, February 5). Roll call vote on Article I, impeachment of President Donald John Trump. U.S. Senate Roll Call Votes, 116th Congress, 2nd Session.

Wisconsin Consumer Act, Wis. Stat. ch. 427 (1971).

ABSTRACT

This work is not abstract. This research paper presents a novel systemic financial contagion risk that I have already reported directly to the SEC under oath. No prior academic work, regulatory framework, or financial stability analysis has identified or modeled the specific risk that systematic noncompliance with mandatory federal and state medical debt disclosure requirements—embedded at the template level in the revenue cycle management platforms that process substantially all US healthcare billing—could propagate through securitization structures to create systemic “phantom capital.”

This paper introduces nine novel terms:

1. retroactive statutory nullification (the mechanism by which a statute renders an entire class of previously executed contracts void by operation of law, with no cure provision and no prospective-only limitation—a term that does not appear in any court opinion, law review article, legal treatise, or financial regulatory publication);

2. phantom debt (individual medical debt obligations void or carrying undisclosed contingent liabilities due to statutory noncompliance);

3. inverse obligation (the instrument-level characterization of a phantom debt as an obligation that generates liabilities in the opposite direction from the cash flows the balance sheet records);

4. phantom receivables (the unsortable aggregation of phantom debt across provider balance sheets);

5. phantom capital (the systemic result when phantom receivables enter the global financial system through layers of leverage);

6. liability inversion (the balance sheet transformation where a receivable becomes a net liability);

7. leveraged contingent liability inversion (the amplification of liability inversion through the leverage chain);

8. structural information absence (the condition where sorting information was never generated at origination); and

9. distributed compliance audit (the methodology by which frontier artificial intelligence enables any individual to systematically analyze provider-issued billing documents against applicable federal and state statutes, transforming consumer protection enforcement from a centralized regulatory function into a massively parallel, individually replicable process).

The underlying noncompliance is national in scope: the FDCPA and Regulation F apply in all fifty states, and California’s SB 1061 renders noncompliant contracts void with no cure provision. This literature review across ten domains confirms that the phantom capital framework identifies a genuinely novel category of systemic financial risk. No scholar, regulator, or rating agency has connected these domains to identify what happens when an entire class of consumer debt is rendered legally compromised and those compromised receivables have already been absorbed into the global financial system through approximately 100–720x hidden leverage. The author holds no financial position in any entity discussed herein, has filed three SEC whistleblower complaints under the Dodd-Frank Act (SEC TCR #17729-304-613-692, SEC TCR #17729-613-590-548, and SEC TCR #17732-878-226-191, which includes this Substack post in its totality). I have thrice voluntarily requested that the SEC conduct trading surveillance on my own accounts to independently verify these representations. A supplementary literature review across the full history of U.S. securities whistleblowing confirms that this act is without documented precedent to do once, let alone three times.

The federal regulatory apparatus has been warned.

Introduction

The medical debt collection system in the United States operates under a layered regulatory architecture. At the federal level, the Fair Debt Collection Practices Act (15 U.S.C. § 1692 et seq.) and its implementing regulation, Regulation F (12 C.F.R. § 1006.34), impose mandatory disclosure requirements on every communication used to collect a debt—including the right to dispute, consequences of non-dispute, availability of verification, identification of the original creditor, and a compliant tear-off dispute form. These requirements apply in all fifty states. They are not optional. They are not subject to substantial compliance. A collection communication either contains the mandatory disclosures or it violates federal law.

Layered on top of this federal baseline, every state maintains its own consumer protection statutes governing debt collection, medical billing, and unfair or deceptive trade practices. Wisconsin has the Wisconsin Consumer Act and the Deceptive Trade Practices Act. Texas has the Texas Debt Collection Act and the DTPA. New York has General Business Law § 349. California has the Rosenthal Fair Debt Collection Practices Act, the Unfair Competition Law, and—most consequentially—SB 1061, signed by Governor Newsom on September 24, 2024, which added Civil Code § 1785.27. The statutory remedies vary by jurisdiction: some states provide for statutory damages per violation, some allow treble damages, some create private rights of action with fee-shifting, and some—like California under SB 1061—render noncompliant medical debt contracts void and unenforceable with no cure provision. But the underlying compliance obligation is universal.

The author of this paper discovered—through direct enforcement activity against his own medical debt across multiple providers and multiple states—that healthcare providers and their billing platforms had not built their templates to comply with these requirements. The noncompliance is not an isolated processing error at a single facility. It is embedded in the billing software itself—in the templates generated by electronic health record and revenue cycle management platforms that serve tens of thousands of providers nationwide. Across every provider examined—in California, in Wisconsin, in facilities operated by health systems spanning thirty or more states—the same pattern held: template-level omission of federally and state-mandated disclosures.

What follows from that observation requires nine new terms—each describing a distinct stage or mechanism of the same phenomenon as it scales from a single patient’s billing statement to the global financial system.

Retroactive Statutory Nullification

Retroactive statutory nullification is the mechanism by which a statute renders an entire class of previously executed contracts void by operation of law, with retroactive effect and no cure provision. California’s SB 1061 (Civil Code § 1785.27) is the paradigmatic example: any medical debt contract entered on or after July 1, 2025 that does not include the statutorily prescribed disclosure language is “void and unenforceable” from the moment of execution. The statute does not impair the contract, does not create a right to rescind, and does not provide a grace period for compliance. It nullifies. The contract never existed as a legal instrument. This is not prospective regulation—it is retroactive annihilation of obligations that the parties believed were valid at the time of formation.

To understand why a state legislature would take this extraordinary step, one must confront what the medical debt asset class actually represents. It is not an abstraction. It is not a line item on a balance sheet that materialized from a neutral commercial transaction between equal parties. Medical debt is the systematic, industrialized repackaging of human suffering into a tradeable financial instrument. It is generated at the moment a human being—sick, injured, frightened, in pain, facing mortality or disability or the medical consequences of conditions they did not choose—is presented with a billing document they did not negotiate, could not have negotiated, and in most cases cannot meaningfully evaluate. The patient signs because the alternative is not receiving care. The patient pays because the alternative is collection, credit destruction, and financial ruin. The entire apparatus of medical billing in the United States is built on the structural exploitation of a power asymmetry so extreme that no other commercial context tolerates it: the provider holds the patient’s health, and the patient has no bargaining position whatsoever.

The billing templates that generate medical debt were not designed to comply with consumer protection statutes. They were designed to extract revenue. The omission of mandatory federal and state disclosure language—the right to dispute, the availability of verification, the consequences of non-dispute, the protections against credit reporting—is not an oversight. It is an architecture. These are the disclosures that would inform patients of their rights, that would empower them to challenge charges, that would create a record of statutory compliance. Their absence is functional: it maximizes the probability that the patient pays without question, without dispute, without awareness of the legal framework that exists to protect them. The platforms that generate these billing communications—athenahealth (athenahealth, Inc., 2024), Epic, Oracle Health, and others—were built to optimize revenue cycle performance, not statutory compliance. When Regulation F and state consumer protection statutes imposed mandatory disclosure requirements, the platforms did not rebuild their templates. The templates predate the statutes. The omissions are omissions of design.

SB 1061 is the legislative response to this reality. It does not merely regulate medical debt. It declares that a medical debt contract formed without the mandatory disclosures—the disclosures that would have informed the patient of their rights at the moment they were most vulnerable—never legally existed. The statute is radical in its remedy precisely because the underlying practice is radical in its exploitation. The California legislature looked at a system that generates $220 billion in financial obligations from the suffering of a hundred million Americans (AMA Council on Medical Service, 2023; KFF & Peterson Health System Tracker, 2023)—a system that assumes patients will never read the statutes, never assert their rights, never discover that the billing documents violating their rights are the very documents that the financial system treats as performing assets—and the legislature said: if you did not tell the patient what the law requires you to tell them, the debt does not exist. Fifteen to sixteen additional states are reaching the same conclusion (National Consumer Law Center, 2025).

This is the context in which retroactive statutory nullification must be understood. It is not an exotic financial risk that emerged from technical regulatory complexity. It is the predictable consequence of building a $220 billion financial asset class on the assumption that vulnerable human beings would never fight back—and then discovering that the legal instruments underlying that asset class were defective from the moment they were created, because the institutions that created them prioritized extraction over compliance. The medical debt asset class does not deserve a haircut. It does not deserve a restructuring. It does not deserve a bailout. It deserves to be shattered into a trillion pieces and never reconstituted, because it was built on the systematic exploitation of patients who were at their lowest and most vulnerable, and the contracts that created it were void from the day they were signed. This paper proves the now-incontrovertible truth that this asset class has poisoned the international financial system through the same leverage structures that amplified the subprime mortgage crisis—except this time, the contamination cannot be sorted from performing assets, the recovery rate is zero, and there is no asset for any government to buy.

The term is entirely novel. An exhaustive search across Google Scholar, SSRN, HeinOnline, SEC EDGAR, the Federal Reserve, BIS, CFPB, Congress.gov, GovInfo, and the open web returned zero results for the exact phrase “retroactive statutory nullification” as a continuous three-word term. All variant phrasings—“statutory nullification retroactive,” “retroactive nullification by statute,” “retroactive nullification of contracts,” “statutory retroactive nullification”—are equally absent. The component concepts are individually well-established: “retroactive legislation” is extensively litigated (Landgraf v. USI Film Products, 511 U.S. 244, 1994; Bowen v. Georgetown University Hospital, 488 U.S. 204, 1988), “nullification” has deep constitutional roots (Jefferson’s Kentucky Resolutions, 1798; Cooper v. Aaron, 1958), and “void ab initio” is centuries-old contract doctrine. But no prior work has combined them into a single term of art describing the specific phenomenon where a state legislature renders an entire class of consumer debt void by operation of statute with retroactive effect. Legal scholarship uses “retroactive repeal,” “retroactive abrogation,” “statutory override,” and “legislative displacement”—none of which captures the specific mechanism of complete voiding (as opposed to amendment or limitation) by legislative act (as opposed to judicial determination) with backward-reaching temporal effect (as opposed to prospective-only application). The terminological gap is not accidental: no prior statute has done to a financial asset class what SB 1061 does to medical debt, and no scholar has needed the term because the phenomenon it describes had not occurred.

Phantom Debt

Phantom debt is a medical debt obligation that appears real on a provider’s balance sheet but is legally impaired, unenforceable, or void at formation because the underlying contract or collection communication fails to comply with mandatory federal and state disclosure requirements. Under California’s SB 1061 (Civil Code § 1785.27), a noncompliant medical debt contract entered after July 1, 2025 is void and unenforceable. Under the FDCPA and Regulation F, noncompliant collection communications create per-violation statutory liability that can exceed the face value of the underlying debt many times over.

The critical insight is that a phantom debt does not simply have a value of zero. It has a negative value. The balance sheet item does not merely disappear—it inverts. Every collection communication sent on a phantom debt generates additional per-violation statutory liability under the FDCPA ($1,000 per violation), Regulation F, and applicable state statutes. Every payment extracted is subject to disgorgement. A $156.20 medical balance, once it becomes phantom debt, can generate tens or hundreds of thousands of dollars in statutory liability for every single instance of noncompliant routine automated collection activity. The asset on the balance sheet is not worth zero. It is an undisclosed contingent liability masquerading as a receivable.

A phantom debt is not impaired debt in the traditional sense. The distinction is ontological, not economic. A defaulted debt is a real obligation—it retains some recovery value, typically 60–70% for senior secured claims (Moody’s Investors Service, 2023). A fraudulent debt is voidable—it exists until someone acts to rescind it. A phantom debt is an inverse obligation that occupies space on a balance sheet as though it were an asset while simultaneously generating liabilities with every automated collection action taken against it. In California, the legal instrument was never formed. In every other state, the obligation carries undisclosed contingent liabilities that dwarf the face value.

The term “phantom debt” is novel in its proposed context. An exhaustive search across Google Scholar, SSRN, SEC EDGAR, Federal Reserve, BIS, CFPB, FTC, and the open web confirmed that the exact phrase appears in three prior contexts, none of which describes the phenomenon identified here. The CFPB and FTC use “phantom debt” to describe fabricated obligations in consumer fraud enforcement actions (CFPB, 2015; FTC, 2019). Wells Fargo Economics used “Phantom Debt” to describe unreported Buy Now, Pay Later obligations creating a macroeconomic measurement blind spot (Quinlan & Grein, 2023). Wikipedia treats the term as synonymous with “zombie debt”—time-barred or discharged debts still pursued for collection (see Sobol, 2014). The definition proposed here occupies a distinct fourth category: medical debt obligations that are void or carry undisclosed contingent liabilities exceeding their face value due to statutory noncompliance at the template level. No prior use describes debt that is legally impaired at formation by the billing platform’s own architectural noncompliance with mandatory disclosure statutes. The distinction is that phantom debt in this usage arises not from fraud, not from measurement failure, and not from the passage of time, but from the legal architecture of the obligation itself.

The sub-term inverse obligation—introduced above to characterize phantom debt at the instrument level—is entirely novel. The exact phrase returns zero results across Google Scholar, SSRN, NBER, SEC EDGAR, and all web sources searched. No financial taxonomy, accounting standard, or legal treatise contains a category for an obligation that generates liabilities in the opposite direction from the cash flows the balance sheet records. The closest existing concepts—“negative equity” (market value decline below secured debt), “toxic assets” (assets whose value has deteriorated), and “stranded assets” (assets that lose value due to regulatory change)—all describe instruments whose value declines. An inverse obligation does not decline in value. It inverts: the instrument actively generates statutory liabilities with every collection action, every billing cycle, every credit report. It is a fundamentally different category of financial instrument that has not been named because the statutory conditions that create it had not previously existed at scale.

Phantom Receivables

When phantom debt is aggregated across a healthcare provider’s patient population, it produces phantom receivables—line items in accounts receivable that represent expected future cash flows from contracts that are legally impaired or nonexistent but cannot be identified as such within the provider’s financial reporting system. Phantom receivables are indistinguishable from performing receivables because the information needed to differentiate them was never generated at origination. Billing templates were not tagged to record which version of disclosure language each patient’s contract contained. No party possesses this sorting information. It does not exist in any database.

This is the sorting problem, and it is unsolvable through ordinary due diligence. Every information asymmetry framework in the financial economics literature—from Akerlof’s (1970) lemons to Gorton and Holmström’s (2020) information-insensitive debt—assumes that sorting information exists somewhere in the system. Phantom receivables present a categorically different condition: the sorting information was never created.

The contamination is likely correlated rather than randomly distributed—and the correlation mechanism is the revenue cycle management (RCM) industry. The major platforms—athenahealth, Epic, Cerner (now Oracle Health), eClinicalWorks, Veradigm—were built before the post-2022 wave of mandatory disclosure requirements. athenahealth alone processes over 315 million claims per year across 170,000+ providers with approximately 17% EHR market share (athenahealth, Inc., 2024). If its templates are noncompliant, the contamination is systematically correlated across its entire provider network.

The liability inversion compounds at the receivables level. The provider has claims exposure to patients, and the provider has corresponding claims against the RCM vendor. athenahealth’s 170,000+ providers are not just its customers—they are its potential plaintiffs. athenahealth was acquired by Bain Capital and Hellman & Friedman for $17 billion in 2022 (Bain Capital, 2022). The contingent liability from template-level noncompliance is not reflected in that valuation or in the $5.9 billion term loan that financed the acquisition (Moody’s Investors Service, 2022).

The term “phantom receivables” is entirely novel. The exact phrase returns zero results across every database searched—Google Scholar, SSRN, SEC EDGAR, Federal Reserve, BIS, CFPB, and the open web. No academic paper, regulatory document, accounting standard, or financial filing uses this phrase in any context whatsoever. It must be distinguished from “fictitious receivables”—a well-established term in accounting fraud literature (ACFE Fraud Examiners Manual; Journal of Accountancy) describing fabricated receivables from non-existent customers created as part of revenue fraud. Fictitious receivables never had a real underlying transaction. Phantom receivables arise from real transactions—real patients received real medical services—but the contracts governing payment are legally impaired or void. It must also be distinguished from “impaired receivables” under IFRS 9, which addresses expected credit losses through normal commercial risk, not legal voidness from statutory noncompliance. The receivable looks real because the transaction was real. But the legal instrument that would make the receivable enforceable was defective at formation.

Phantom Capital

Phantom capital is what phantom receivables become when they enter the financial system through layers of leverage. The United States carries approximately $220 billion in outstanding medical debt (KFF & Peterson Health System Tracker, 2023). Healthcare and pharmaceutical companies constitute roughly 25.5% of all leveraged loans (ZAIS Group, 2024). Medical receivables enter indirectly: patient revenue streams service corporate debt, syndicated into leveraged loans, purchased by CLOs leveraging equity approximately 10–11x (Office of Financial Research [OFR], 2024), held by life insurers at asset-to-equity ratios approaching 12x (Federal Reserve, 2025), banks, and retail investors through CLO ETFs that grew from $120 million in 2020 to over $19 billion by late 2024 (Bloomberg, 2024). Medical credit card products—CareCredit, Alphaeon Credit—are potentially securitized within general credit card ABS pools.

At every stage, the system assumes the underlying medical debt contracts are legally valid. The theoretical amplification: 6x × 10x × 12x = 720x at the extreme. When the asset at the base is legally nonexistent—recovery is zero—the amplification operates on a total loss rather than the 30–40% loss-given-default that every model assumes (Moody’s Investors Service, 2023).

This is phantom capital: balance sheet value built on leverage applied to assets that do not legally exist or that are liabilities masquerading as receivables. The leverage does not merely amplify a loss. It amplifies an inversion. The term is novel in financial economics. An exhaustive search across Google Scholar, SSRN, SEC EDGAR, Federal Reserve, BIS, and the open web returned zero results for “phantom capital” in the context of leveraged balance sheet value built on legally nonexistent or compromised assets. The term must be distinguished from Marx’s (1894) “fictitious capital” (Capital, Vol. III; see also Hilferding, 1910; Durand, 2017), a well-established concept in political economy describing all financial claims—stocks, bonds, derivatives—as “fictitious” because they represent titles to future surplus value rather than real productive capital. Marx’s concept applies to all financial securities, even perfectly legitimate ones. It must also be distinguished from the IMF’s “Phantom FDI” (Damgaard et al., 2019), which describes $15 trillion in foreign direct investment passing through empty corporate shells in tax havens—a related use of “phantom” in a financial context but addressing cross-border tax arbitrage, not leveraged balance sheet value on void assets. Phantom capital in the present usage is narrower and forensic: it targets specifically the leveraged amplification of legally compromised or nonexistent assets within securitization structures.

Liability Inversion

Liability inversion is the balance sheet transformation that occurs when a receivable becomes a net liability generating cash outflow obligations that exceed its face value. Driven by per-violation statutory damages under the FDCPA ($1,000 per violation), Regulation F, and state consumer protection statutes, compounded by disgorgement obligations and litigation exposure. The receivable does not merely lose value. It flips.

The term “liability inversion” is novel. An exhaustive search across Google Scholar, SSRN, SEC EDGAR, Federal Reserve, BIS, and the open web returned no use of this phrase as a defined concept in any academic, regulatory, or financial publication. The only documented instance is in Accountfy accounting software documentation (2021), where “liability inversion” refers to a trivial UI feature that multiplies liability figures by negative one for chart display—entirely unrelated. The underlying phenomenon—statutory per-violation damages creating potentially ruinous aggregate liability—is recognized but unnamed. Scheuerman (2011) discussed how per-violation damages become a “multiplier of liability” that can be firm-threatening when aggregated in class actions. But no source frames this as a balance sheet transformation where an asset becomes a net liability, and none uses the term. It must be distinguished from “negative equity” or “underwater assets” (market value decline below secured debt, driven by economic forces), from “flip clauses” in securitization (provisions changing payment priority upon trigger events), and from corporate “tax inversions” (relocating corporate domicile to lower-tax jurisdictions). Liability inversion is categorically different: it is a statutory mechanism, not a market mechanism, and it converts the nature of the balance sheet item rather than merely reducing its value.

Leveraged Contingent Liability Inversion

Leveraged contingent liability inversion describes what happens when liability inversion propagates through the leverage chain. The leverage chain does not merely transmit a loss downward. It generates a litigation chain running upward—from investors back toward origination. The CLO faces deterioration no rating model anticipated because the model assumed the revenue was real (Benmelech & Dlugosz, 2009). The life insurer discovers capital relief was predicated on assets generating liabilities (Carlino et al., 2025). Each layer multiplies not expected return but the magnitude of an undisclosed contingent liability.

This term is entirely novel—the most definitively novel of the nine terms introduced in this paper. The exact phrase “leveraged contingent liability inversion” returns zero results across every database searched: Google Scholar, SSRN, NBER, SEC EDGAR, Federal Reserve, BIS, and the open web. No close variant exists either: “contingent liability inversion” returns zero results; “leveraged liability inversion” returns zero results. The component concepts each exist independently in well-established literature—“contingent liability” is codified in IAS 37 and ASC 450; “leverage” is fundamental to finance (Basel III frameworks); “inversion” appears in yield curve and ETF contexts—but their combination into a single analytical construct describing the phenomenon where leverage amplifies a contingent liability that inverts an asset’s value has never been articulated. The closest related work—Esty (1998) on contingent liability and bank risk-taking, Grossman (2013) on contingent liability and leveraged British banks, and the Bank of England’s discussion of “contingent ‘hidden’ leverage” during the 2022 UK LDI/gilt crisis (Breeden, 2022)—addresses at most two of the three component concepts without combining them into the specific phenomenon described here.

Structural Information Absence

Structural information absence is the condition where the information needed to sort phantom from performing receivables was never generated at origination. This is fundamentally different from every category in the information economics literature. Akerlof’s (1970) adverse selection assumes the seller possesses information. Stiglitz and Weiss (1981) assume borrowers know their prospects. The originate-to-distribute literature (Keys et al., 2010; Mian & Sufi, 2009) frames screening failure as moral hazard. Gorton and Holmström (2015, 2020) describe a rational choice to suppress information demand. Structural information absence is categorically different: the screening mechanism itself was absent. The standard taxonomy—hidden (adverse selection), unobservable (moral hazard), costly to produce (Stigler, 1961), strategically suppressed (Gorton & Holmström, 2020)—has no category for information that was never generated. Crémer and Khalil (1992), Hart and Moore (1990), and Knightian uncertainty (Knight, 1921) each address different phenomena.

A precise analogy from computer science clarifies the distinction. In database systems, a null value is a defined state: the field exists in the schema, the system expects a value, and null is the explicit representation of “no value present.” A null is a known unknown—the system knows the field is empty and can query for its absence. Every information problem in the existing economics literature is analogous to a null: the information field exists in the conceptual schema of the transaction, and the problem is that the value is missing, hidden, misrepresented, or strategically withheld. Structural information absence is not a null. It is the condition where the field itself does not exist—where the database schema was never built to include a column for “compliance status of billing template at time of origination.” No party can query for the value because no party’s system contains the field. The information cannot be retrieved, imputed, or reconstructed, because the measurement instrument that would have generated it was never deployed. This is not a missing value in an existing data structure. It is the absence of the data structure itself.

This term and concept are entirely novel. The phrase “structural information absence” returns zero results across Google Scholar, SSRN, NBER, and all web sources searched. The novelty lies not merely in the phrase but in classifying the phenomenon as a distinct information-theoretic category rather than merely an incentive failure.

Distributed Compliance Audit

A distributed compliance audit is the process by which an individual uses frontier artificial intelligence tools to systematically analyze provider-issued billing statements, collection communications, and medical debt contracts against the full text of applicable federal and state statutes—including the FDCPA (15 U.S.C. § 1692 et seq.), Regulation F (12 C.F.R. § 1006.34), SB 1061 (Cal. Civ. Code § 1785.27), the Wisconsin Consumer Act, state DTPA statutes, and their counterparts in all fifty states—to produce a complete compliance determination with per-violation damages calculations and enforcement-ready regulatory filings. The term describes both the methodology and the structural consequence: when any patient with access to frontier AI can conduct the same level of rigorous statutory analysis that previously required specialized consumer protection counsel, the enforcement function is no longer centralized in regulatory agencies or contingency-fee law firms. It is distributed across the entire patient population.

This is the mechanism that makes the phantom capital crisis an immediate rather than theoretical risk. Without the distributed compliance audit, the template-level noncompliance identified in this paper would remain latent—present in the system but undiscovered by the patients whose rights are being violated. The traditional enforcement bottleneck—the cost and complexity of retaining counsel, researching applicable statutes, calculating damages, and drafting complaints—suppresses the assertion rate of valid consumer protection claims to a fraction of eligible patients. Frontier AI eliminates that bottleneck entirely. A patient can upload a billing statement to a frontier AI system and receive, within a single session, a complete analysis identifying every statutory violation, calculating per-violation damages under every applicable federal and state statute, and producing finished regulatory complaint documents addressed to every relevant agency. The author of this paper developed this methodology through direct enforcement activity against his own medical debt and has published the framework publicly (Ibiza, 2026a, 2026b). The methodology is fully replicable. It requires no legal training, no specialized software, and no institutional support—only a billing statement and access to a frontier AI platform.

The systemic implications are profound. The existing financial stability literature assumes that consumer protection enforcement operates through two channels: public enforcement by regulatory agencies (CFPB, FTC, state AGs) and private enforcement by plaintiffs’ attorneys (typically on contingency in class actions). Both channels are capacity-constrained—limited by agency budgets, staffing, political priorities, and the economics of contingency-fee litigation. The distributed compliance audit creates a third channel: individual enforcement at scale, enabled by AI, with no capacity constraint. Every patient who has ever received a noncompliant billing communication is a potential enforcer. The barrier to enforcement drops from “retain a consumer protection attorney” to “upload a PDF.” The traditional assumption that only a small fraction of statutory violations will ever be enforced—the assumption on which the economic viability of noncompliant billing practices depends—is invalidated.

The term “distributed compliance audit” is entirely novel. An exhaustive search across Google Scholar, SSRN, HeinOnline, SEC EDGAR, Federal Reserve, BIS, CFPB, FTC, Congress.gov, and the open web returned zero results for the exact phrase. The closest existing concepts are “crowdsourced regulation” (Oei & Ring, 2018), “regulatory technology” or “RegTech” (Arner et al., 2017), and the private attorney general doctrine (Newman v. Piggie Park Enterprises, 1968). None describes the specific phenomenon of AI-enabled, individually replicable, massively parallel consumer protection enforcement conducted by the rights-holders themselves. “Crowdsourced regulation” describes voluntary citizen participation in regulatory monitoring (e.g., Waze for road conditions). “RegTech” describes technology used by regulated entities to comply with obligations. The distributed compliance audit is categorically different: it is technology used by the people the regulations were designed to protect, to discover violations by the entities the regulations were designed to constrain, and to produce enforcement-ready filings that those entities cannot ignore because the statutory violations are binary—the mandatory language is either present or it is not. The methodology cannot be suppressed, because the statutes are public, the billing documents are in the patients’ possession, and the AI tools are commercially available. Once published, it is permanently replicable by anyone.

The Scale of the Problem

The template-level noncompliance is not confined to a single state. Regulation F and the FDCPA apply in all fifty states. California’s SB 1061 represents the sharpest consequence, but 15–16 additional states have enacted or are pursuing similar medical debt protection legislation (National Consumer Law Center, 2025). The $220 billion medical debt asset class is contaminated with phantom receivables that cannot be sorted from performing assets. The contamination propagates upward through layers of leverage: provider balance sheets, asset-based lending, corporate credit ratings, CLO tranches, institutional portfolios, counterparty exposure, derivatives. No Federal Reserve stress test models this scenario. No rating agency has published commentary on SB 1061. No Basel capital framework contains a variable for “percentage of receivables retroactively rendered legally void.” The risk is not merely unpriced—it is uncategorized.

Purpose of This Review

This literature review examines ten distinct domains to confirm that the framework described above—retroactive statutory nullification, phantom debt, inverse obligation, phantom receivables, phantom capital, liability inversion, leveraged contingent liability inversion, structural information absence, and distributed compliance audit—identifies a genuinely novel category of systemic financial risk. Each domain is well-developed in isolation. None has been connected to the others through the specific mechanism of retroactive statutory nullification or systematic statutory noncompliance in consumer debt.

1. The post-2008 securitization literature mapped every risk except retroactive statutory nullification

The academic response to the 2008 financial crisis produced a comprehensive taxonomy of securitization risk—yet that taxonomy contains no category for retroactive statutory nullification—contracts voided by operation of statute. Gorton’s (2010) foundational work, Slapped by the Invisible Hand, and the information-insensitive debt framework developed with Holmström (Dang et al., 2020), established that opacity in debt markets is functional: short-term debt is designed so that no party has incentive to produce costly private information about it.

Gorton’s entire framework assumes the underlying contracts are legally valid—the information sensitivity concerns economic valuation, not legal existence.

Coval et al. (2009) demonstrated that senior tranches of CDOs function as economic catastrophe bonds whose risk is highly systematic, and that structuring amplifies errors in risk estimation—but frames the “error” exclusively as parameter mis-estimation, not legal nonexistence. Levitin and Wachter (2012, 2020) identified the crisis as supply-side mispricing. Adrian and Ashcraft (2010, 2012) documented shadow banking risks—all economic, none legal. Acharya et al. (2017) introduced Systemic Expected Shortfall; Acharya et al. (2013) showed securitization concentrated risk. Neither contemplated legally invalid assets.

The complete risk taxonomy encompasses credit risk, fraud risk, model risk, correlation risk, liquidity risk, agency risk, regulatory arbitrage risk, and opacity risk. Retroactive statutory nullification risk—the possibility that a statutory change could render an entire class of consumer debt void ab initio—appears nowhere.

2. Information asymmetry literature describes unknowns, not unknowables

Keys et al. (2010) demonstrated that securitization reduced screening incentives—loans more easily securitized defaulted 10–25% more. Rajan et al. (2015) showed lenders strategically ignored soft information. Ashcraft and Schuermann (2008) identified seven informational frictions. Akerlof’s (1970) framework assumes the seller possesses sorting information the buyer lacks. Every paper assumes the sorting information exists. The phantom capital scenario presents structural information absence: information that was never generated at origination.

Gorton and Holmström’s (2020) insight that opacity is functional acquires a disturbing new dimension: when purposeful opacity is paired with a statutory regime that voids contracts lacking specific language, the opacity that normally facilitates trading becomes the vector through which void contracts enter the financial system. The functional feature becomes the vulnerability.

3. CLO and leverage propagation literature quantifies amplification but assumes valid assets

Cordell et al. (forthcoming) documented typical CLOs hold 150–250 loans with approximately 90% leverage. Benmelech and Dlugosz (2009) showed B-rated pools transform into ~73% AAA-rated liabilities. Griffin and Nickerson (2023) found CLOs appear considerably riskier than current ratings suggest. Foley-Fisher et al. (2020) documented how CLO AAA tranches become information-sensitive during stress. Carlino et al. (2025) documented life insurer regulatory arbitrage cutting capital charges by two-thirds. The Financial Stability Board (2023) identified “hidden leverage” across non-bank financial intermediation. The CLO market stands at $1.3–1.4 trillion outstanding (Loan Syndications and Trading Association [LSTA], 2024), with CLOs purchasing 61% of all new leveraged loans (LCD/Pitchbook, 2024) and holding approximately 64–70% of institutional leveraged loan outstandings.

Brunnermeier and Pedersen (2009) formalized the loss spiral and margin spiral. Brunnermeier (2009) documented $120 billion in mortgage losses amplifying into over $8 trillion—roughly 60–70x. Geanakoplos (2010) developed leverage cycle theory.

No paper analyzes what happens when a receivable at the base is void, recovery is zero, or the asset class is correlated, collapsing diversification precisely as Coval et al. (2009) warned.

4. Medical debt is a $220 billion exposure the financial stability literature has ignored

The CFPB’s (2022) report established $88 billion on credit records, 58% of all third-party collection tradelines. The KFF/Peterson tracker (2023) estimated “at least $220 billion.” Fulford and Wilson (2025) found $194 billion in active collection, 36% of US households affected. The AMA Council on Medical Service (2023) reported approximately 100 million Americans affected.

Medical debt enters the financial system indirectly: healthcare leveraged loans, AR-based lending, medical credit card products (CareCredit, Alphaeon Credit), and whole-loan portfolio sales. Healthcare and pharmaceuticals constitute approximately 25.5% of all leveraged loans, with ~56.9% of loans in these sectors rated below B-flat (ZAIS Group, 2024).

athenahealth—The Transmission Mechanism

athenahealth is the transmission mechanism. Its revenue cycle management platform generates the billing templates, statement generators, and collection communication workflows that create de facto standardization across its provider network. If athenahealth’s templates are noncompliant with Regulation F and state disclosure requirements—and the author’s direct enforcement activity confirms that they are—then every provider on the platform generates noncompliant billing communications using the same deficient architecture. The contamination is not idiosyncratic. It is systematically correlated at the platform level. athenahealth is one platform among several—Epic, Cerner (Oracle Health), eClinicalWorks, Veradigm, and others collectively manage billing for the nation’s hospitals, surgery centers, and physician practices. These platforms were built before the post-2022 wave of mandatory disclosure requirements.

Surgery Partners, Inc. (NASDAQ: SGRY)—The Leveraged Loan Exposure

Surgery Partners, Inc. (NASDAQ: SGRY), also a Bain Capital portfolio company (~39% ownership; Bain Capital, 2024), illustrates what sits downstream of the transmission mechanism. SGRY’s $1.4 billion broadly syndicated term loan is precisely the type of floating-rate, sub-investment-grade debt that populates CLO portfolios (Surgery Partners, Inc., 2025). The revenue servicing this debt derives from medical receivables—insurance reimbursements and patient self-pay obligations generated through billing platforms including athenahealth. If SB 1061 renders a meaningful fraction of patient obligations at SGRY’s California facilities void, the revenue supporting the leveraged loan is impaired, yet neither the CLO model nor the corporate credit rating captures this specific risk pathway. Bain Capital sits at the nexus of both the template vector (athenahealth) and the leveraged exposure (SGRY)—a vertically integrated position in which the same private equity firm controls the platform generating phantom debt and the healthcare company whose balance sheet absorbs it.

No academic paper, CFPB report, FSOC publication, or financial stability analysis has examined medical debt through the lens of systemic financial risk via securitization channels or retroactive statutory nullification.

5. California SB 1061 creates a void-contract trigger with no cure and no precedent in financial risk analysis

SB 1061, signed September 24, 2024, added Civil Code § 1785.27 with two void-and-unenforceable triggers. Section (b) renders a medical debt void if someone knowingly furnishes information about it to a CRA. Section (c)(2) provides that any contract entered on or after July 1, 2025 without the prescribed disclosure language is “void and unenforceable”—with no cure provision. The required language must inform the patient that the debt holder is prohibited from furnishing information to CRAs and that knowing violation renders the debt void.

The federal preemption landscape adds uncertainty. The Biden-era CFPB rule was vacated in July 2025. The Trump-era CFPB claimed FCRA preempts state bans. California AG Bonta reaffirmed state law. The First Circuit’s decision in Consumer Data Industry Association v. Frey (2022) supports non-preemption. The question remains unresolved—a Schrödinger’s-cat condition for medical debt receivables.

No rating agency has published any commentary on SB 1061. No financial stability body has assessed its implications. No stress test includes it as a scenario.

6. The void/voidable distinction has never been analyzed in financial risk literature

An exhaustive search returned no results directly analyzing the financial system consequences of void ab initio contracts in securitization.

The Georgia Fair Lending Act Precedent (2002–2003)

Georgia enacted the GFLA in October 2002 with assignee liability provisions. In January 2003, S&P announced it would no longer rate MBS containing Georgia mortgages; Moody’s and Fitch followed, freezing the secondary mortgage market. The legislature capitulated in March 2003, passing HB 200. But the GFLA involved voidable contracts and damages liability, not void contracts.

Gerding (2009) is the single paper most proximate to the phantom capital thesis—but analyzed regulatory consumer protection, not the legal validity of contracts themselves.

7. Regulatory frameworks contain no category for legal nullification risk

The Basel Committee classifies legal risk as a sub-category of operational risk. The Basel securitization framework (revised January 2018) assumes valid, enforceable claims. There is no scenario variable for “X% of receivables are retroactively rendered legally void.”

Federal Reserve stress tests model macroeconomic scenarios. No scenario includes a legislative event rendering a class of debt unenforceable. FSOC’s Analytic Framework has never considered consumer debt retroactive statutory nullification as a systemic risk event.

The fundamental gap: there is no “legislative event” variable in any stress test, capital model, or systemic risk framework.

8. Rating agencies analyze structural isolation but never question whether the asset itself is real

All three major agencies take corporate credit ratings as given CLO inputs without “looking through” to the legal enforceability of revenue streams.

DBRS Morningstar, the most active rater of healthcare receivables securitizations, focuses on provider credit quality, not legal enforceability. No rating agency has published commentary on SB 1061. Fender and Kiff (2004), Hill (2010), and Rösch and Scheule (2011) documented rating agency failures—but never contemplated legislative voiding of an entire asset class.

9. Private enforcement as a financial stability vector has never been examined

No academic paper connects private enforcement of consumer protection statutes to systemic financial consequences. Berger et al. (2022) noted that even public enforcement’s relationship to systemic risk was uninvestigated prior to their paper. The Georgia GFLA episode demonstrates financial system sensitivity to private enforcement vectors, yet was treated exclusively as a lobbying story. This section examines the legal doctrine that enables private enforcement, the technological development that removes the practical barriers to its exercise, and the systemic implications of their convergence.

The Private Attorney General Doctrine

The legal foundation for private consumer protection enforcement is the private attorney general doctrine, established by the Supreme Court in Newman v. Piggie Park Enterprises, Inc., 390 U.S. 400 (1968). The doctrine recognizes that when Congress creates statutory rights enforceable through private litigation with fee-shifting provisions, the private plaintiff serves a public enforcement function—vindicating not merely their own rights but the rights of all similarly situated persons and the public interest in statutory compliance. Fee-shifting statutes—including the FDCPA (15 U.S.C. § 1692k), state consumer protection acts with attorney’s fees provisions, and the private right of action under California’s Unfair Competition Law (Cal. Bus. & Prof. Code § 17200 et seq.)—are designed to incentivize private enforcement by ensuring that successful plaintiffs can recover the costs of litigation even when individual damages are small.

The doctrine has deep roots in American law. Alyeska Pipeline Service Co. v. Wilderness Society, 421 U.S. 240 (1975), constrained judicially created fee-shifting but prompted Congress to enact the Civil Rights Attorney’s Fees Awards Act of 1976 (42 U.S.C. § 1988), codifying the principle that private enforcement of statutory rights warrants fee recovery. Every subsequent consumer protection statute with a fee-shifting provision builds on this framework. The FDCPA’s damages structure—$1,000 per violation in individual actions, plus actual damages, plus attorney’s fees—is explicitly designed to make private enforcement economically viable for claims that would otherwise be too small to litigate. The entire architecture of American consumer protection enforcement assumes and depends upon private plaintiffs exercising this function.

Yet the financial stability literature has entirely ignored this enforcement channel. Every analysis of systemic risk in securitization structures—from Coval et al. (2009) through Gorton and Holmström (2020)—models the risk of asset impairment through economic forces: borrower default, credit deterioration, macroeconomic shocks. None models the risk that private enforcement of consumer protection statutes could render the underlying assets legally void. The private attorney general is invisible in the financial stability framework—an enforcement mechanism with the legal authority to trigger asset class destruction that no risk model has been designed to detect.

The Distributed Compliance Audit as Force Multiplier

Historically, the private attorney general doctrine operated under severe practical constraints. The fee-shifting incentive attracted plaintiffs’ attorneys to high-value or class-certifiable claims, but individual consumer protection violations—a missing disclosure on a single billing statement—rarely justified the cost of investigation, statutory research, damages calculation, and complaint drafting. The enforcement bottleneck was not legal but practical: the doctrine gave every patient the right to enforce, but the economics of enforcement limited actual exercise of that right to a small fraction of eligible claims. The financial system’s exposure to private enforcement was therefore bounded by the capacity of the consumer plaintiffs’ bar, and that capacity constraint was implicitly assumed in every institutional risk model that failed to account for private enforcement as a systemic risk vector.

The distributed compliance audit, as defined in the Introduction to this paper, eliminates that capacity constraint. Frontier AI enables any individual to conduct the same level of rigorous statutory analysis that previously required specialized consumer protection counsel—and to do so in a single session, at negligible cost, with no legal training. The practical barrier that limited private attorney general enforcement to a fraction of eligible claims is removed. The legal right that the doctrine conferred on every patient is now matched, for the first time, by a practical capability to exercise it.

The implications for the phantom capital framework are direct. The template-level noncompliance identified in this paper affects every patient billed through every noncompliant RCM platform—potentially tens of millions of individuals. The private attorney general doctrine gives each of those individuals a statutory right to enforce. The distributed compliance audit gives each of them the practical means to discover the violation, calculate the damages, and produce enforcement-ready filings. The convergence of a legal right that has existed since 1968 with a technological capability that emerged in 2024–2025 creates an enforcement surface that is unprecedented in scale: not a class action filed by a single firm, but a massively parallel, individually initiated enforcement campaign that no defendant can settle with a single payment and no court can consolidate into a single proceeding, because each patient’s claim arises from their own billing documents under their own state’s statutes.

No financial stability scholarship has examined the interaction between the private attorney general doctrine and technological force multiplication. The literature on “regulatory technology” or “RegTech” (Arner et al., 2017) examines technology used by regulated entities to comply with obligations—not technology used by rights-holders to discover violations. The literature on “crowdsourced regulation” (Oei & Ring, 2018) examines voluntary citizen participation in monitoring—not AI-enabled statutory enforcement producing binding legal consequences. The intersection of the private attorney general doctrine, frontier AI, and consumer protection enforcement against securitized asset classes is entirely unexamined in any domain of scholarship.

This intersection—private litigation triggering the discovery that securitized assets are legally nonexistent—has no precedent in financial risk scholarship. No one else has ever done this the way I have in the entire history of securities enforcement in the United States of America.

10. The transmission chain from void receivable to systemic risk

The complete transmission chain can now be specified using the literature’s own frameworks.

WHAT THE EXISTING LITERATURE PROVIDES — AND WHAT IT MISSES The existing literature provides every component of the transmission mechanism: Gorton and Holmström (2020) explain why no one asks whether the underlying contracts are valid. Coval et al. (2009) explain why structuring amplifies any error at the base. Brunnermeier and Pedersen (2009) explain how leverage propagates losses into spirals. Carlino et al. (2025) document the regulatory arbitrage that creates hidden leverage in life insurance CLO holdings. The rating agencies model structural isolation without ever questioning receivable validity. The Basel frameworks model credit risk, market risk, and operational risk without a variable for retroactive statutory nullification. FSOC assesses systemic vulnerabilities without examining medical debt. The literature is comprehensive, rigorous, and blind. What it misses is everything this paper identifies—including nine novel terms, none of which existed in any prior academic, regulatory, or financial publication:

1. Retroactive statutory nullification — the trigger mechanism itself, a term that returns zero results across every database searched. No prior statute has done to a financial asset class what SB 1061 does to medical debt, and no scholar has needed the term because the phenomenon had not occurred.

2. Phantom debt — the individual medical debt obligation that is void, unenforceable, or carries undisclosed contingent liabilities exceeding its face value due to statutory noncompliance.

3. Inverse obligation — the instrument-level characterization: a debt that generates liabilities in the opposite direction from the cash flows the balance sheet records.

4. Phantom receivables — the unsortable aggregation of phantom debt across provider balance sheets, contaminated by structural information absence.

5. Structural information absence — the condition where sorting information was never generated at origination. Not a null value in an existing schema—the schema itself does not exist.

6. Phantom capital — the systemic result: balance sheet value built on approximately 100–720x leverage applied to assets that do not legally exist.

7. Liability inversion — the balance sheet transformation where a receivable flips from asset to net liability through accumulating statutory damages.

8. Leveraged contingent liability inversion — the amplification of liability inversion through the leverage chain, where each layer multiplies not a loss but an undisclosed liability masquerading as an asset.

9. Distributed compliance audit — the methodology by which frontier AI enables any individual to systematically analyze billing documents against statutory requirements, transforming consumer protection enforcement from a centralized function into a massively parallel, individually replicable process. The enforcement bottleneck that protected noncompliant billing practices is eliminated.

No scholar, regulator, rating agency, or financial stability body has identified any of this.

No prior work has named any of these phenomena.

This paper is the first to do both.

This presents an immediate regulatory and institutional crisis for all levels of government and enterprise in the United States of America.

11. An Immediate Regulatory Crisis: A Call to Action

This paper is not a theoretical exercise. It is a warning. The systemic risk identified in these pages is not a future contingency that may materialize under adverse conditions. It is a present reality that exists in the financial system right now—undetected by every model, unpriced by every market, and uncategorized by every regulatory framework. The contamination is already embedded in provider balance sheets, already flowing through leveraged loans, already sitting inside CLO tranches held by life insurers, banks, pension funds, and retail investors through CLO ETFs that did not exist five years ago (Bloomberg, 2024). The only question is when the system discovers what it is holding. Not whether. This paper is the discovery.

The 2008 Parallel—and Why This Is Worse

The structural parallel to the 2007–2009 financial crisis is direct. In that crisis, the financial system discovered that mortgage-backed securities—treated as performing assets on the balance sheets of banks, investment firms, and insurance companies—were backed by mortgages that borrowers could not repay. The assets were not worth what the balance sheets said they were worth. The result was a cascading loss of confidence, a freezing of credit markets, and a systemic crisis that required trillions of dollars in government intervention to resolve (Brunnermeier, 2009).

The phantom capital crisis shares the same architecture: assets assumed to have value that do not, leverage amplifying the loss across the financial system, opacity preventing detection until the cascade is underway, and rating agencies that failed to identify underlying quality defects. But the phantom capital crisis is structurally worse in three specific dimensions.

First, there is no floor. In the subprime crisis, the underlying assets lost value because borrowers could not repay—but the houses still existed. The land still existed. Recovery rates on defaulted mortgage-backed securities, while far lower than models predicted, were not zero. Senior secured claims recovered 60–70% (Moody’s Investors Service, 2023). Phantom debt has no recovery rate. The contracts do not fail because patients cannot pay. They fail because the law says they do not exist. There is no house. There is no land. There is no collateral to seize, no borrower to restructure with, no workout to negotiate. The asset at the base of the leverage chain is a legal nullity—and a legal nullity that is simultaneously generating statutory liabilities in the opposite direction from the cash flows every model assumes. The loss-given-default is not 30–40%. It is greater than 100%, because the liability inversion means the “asset” is a net obligation.

Second, there is no asset to buy. In 2008, the crisis could be resolved—at enormous cost—because the government could buy the distressed assets. The Troubled Asset Relief Program (TARP) purchased mortgage-backed securities. The Federal Reserve’s quantitative easing programs absorbed toxic assets onto its balance sheet. The underlying theory was that the assets had value, that the market was mispricing them due to panic, and that a patient buyer with a long time horizon could hold them to recovery. This intervention architecture does not exist for phantom capital. There is no asset to buy. A void contract cannot be purchased at a discount and held to recovery because there is nothing to recover. A phantom receivable cannot be restructured because the legal instrument that would create the obligation was never formed. The entire toolkit of crisis intervention—asset purchases, guarantees, forbearance, restructuring—presupposes that the distressed asset is a real instrument with some recoverable value. Phantom debt is not a distressed asset. It is a non-asset occupying the space where an asset should be. No central bank can buy what does not exist. No TARP can purchase a void contract. No quantitative easing program can absorb a legal nullity onto a balance sheet and wait for it to perform, because performance requires a valid obligation and no valid obligation was ever created.

Third, the contagion mechanism resembles a pandemic more than a financial crisis. In a traditional financial crisis, contagion spreads through counterparty exposure and confidence collapse—institution A’s losses impair institution B’s collateral, triggering margin calls that force institution C to liquidate. The phantom capital contagion operates differently. It spreads through discovery. The contamination is already present in the system. It has been present since the billing templates were built without the mandatory disclosure language. What triggers the crisis is not a new loss event but the recognition that losses have already occurred—that what the system has been counting as assets are in fact liabilities, and that this condition is correlated across every provider on every noncompliant billing platform in every state. The contagion does not spread from institution to institution through balance sheet linkages. It spreads from awareness to awareness—from the first analyst who reads this paper to the first rating agency that examines a billing template to the first CLO manager who asks whether the healthcare company’s revenue includes collections on phantom debt. Each discovery event confirms that the contamination exists everywhere the same noncompliant templates were used. The pandemic analogy is precise: the virus was already circulating before anyone tested for it, and the act of testing did not create the infection—it revealed an infection that had been spreading undetected. Except in a pandemic, the pathogen is biological and the intervention is medical. Here, the pathogen is a legal defect embedded in billing software, and there is no vaccine. The only intervention is compliance—and compliance cannot be applied retroactively to contracts that are already void.

No Intervention Plan Exists

The Federal Reserve’s Financial Stability Report (Board of Governors, 2025) does not mention medical debt, billing template compliance, or retroactive statutory nullification risk. FSOC’s Analytic Framework for Financial Stability does not contain a category for legislative events that void asset classes. No stress test run by any central bank in any country includes a scenario in which a state legislature renders a class of consumer debt legally unenforceable. The Basel Committee’s operational risk framework classifies legal risk under “clients, products, and business practices”—as an individual firm-level loss event, not as systemic asset class destruction. Rating agencies model structural isolation in securitization without ever asking whether the underlying receivable is legally valid.

This means that when the phantom capital contamination is discovered at scale—and this paper, the SEC whistleblower filings, the certified mail campaign, the regulatory complaints, and the distributed compliance audit methodology ensure that discovery is imminent—the regulators responsible for maintaining financial stability will have no playbook, no model, no intervention architecture, and no legal authority designed for this specific category of crisis. They will be improvising with tools built for a different kind of problem.

A Foundational Crisis for Electronic Medical Records in the United States

The phantom capital crisis does not merely expose a gap in financial regulation. It exposes a foundational defect in the national project of electronic medical records—the multi-decade, multi-billion-dollar federal initiative that was supposed to modernize American healthcare infrastructure and that instead created the very correlation mechanism that makes this crisis systemic.

The Health Information Technology for Economic and Clinical Health (HITECH) Act of 2009 authorized approximately $36 billion in federal incentive payments to accelerate the adoption of electronic health record (EHR) systems across the United States. The program succeeded on its own terms: EHR adoption among office-based physicians rose from approximately 48% in 2009 to over 90% by 2021, and nearly all US hospitals now operate on electronic systems. The Office of the National Coordinator for Health Information Technology (ONC) certified these platforms, CMS administered the incentive payments, and the federal government declared the digitization of American healthcare a public policy achievement. The platforms that received certification and whose adoption was subsidized by federal dollars—athenahealth (athenahealth, Inc., 2024), Epic, Cerner (now Oracle Health [NYSE: ORCL]), eClinicalWorks, Veradigm, and others—are the same platforms whose billing templates generate the noncompliant collection communications that create phantom debt.

This is not a coincidence. It is a structural consequence. The HITECH Act and the Meaningful Use program that implemented it focused on clinical data—patient records, interoperability, quality reporting. The revenue cycle management (RCM) modules embedded in these same platforms—the modules that generate billing statements, collection communications, and patient financial responsibility notices—were not subject to the same certification rigor. ONC certification addressed whether the EHR could record a diagnosis, transmit a lab result, or generate a clinical summary. It did not address whether the platform’s billing templates complied with the Fair Debt Collection Practices Act, Regulation F, or state consumer protection statutes. The billing side of the platform was treated as a business function, not a regulated communication. The federal government certified the clinical module and ignored the financial module—and the financial module is the one generating the phantom debt that is now contaminating the global financial system.

The correlation mechanism that makes phantom capital a systemic rather than idiosyncratic risk—the fact that the contamination is platform-level rather than provider-level—is a direct product of the federal EHR adoption mandate. Before HITECH, medical billing was fragmented across thousands of independent systems, paper-based processes, and local configurations. The noncompliance that existed was heterogeneous and uncorrelated. HITECH consolidated American healthcare billing onto a handful of dominant platforms, each serving tens of thousands of providers across all fifty states. The consolidation that the federal government promoted as an efficiency gain is the same consolidation that creates de facto template standardization—and when the standardized template is noncompliant, the noncompliance is systematically correlated across the platform’s entire provider network. The diversification benefits that structured finance relies upon—the assumption that medical receivables from different providers in different states represent independent credit risks—collapse because the receivables were all generated by the same noncompliant software architecture that the federal government subsidized.

The implications extend beyond financial stability. If the billing templates embedded in federally certified EHR platforms are generating void or legally impaired medical debt—and the author’s direct enforcement activity across multiple providers on multiple platforms confirms that they are—then the national EHR infrastructure has a compliance defect that affects every patient interaction processed through these systems. The defect is not clinical. It is financial. But the financial defect undermines the integrity of the entire patient-provider relationship, because every billing communication that omits the mandatory disclosures is a communication that deprives a patient of their statutory rights at the moment they are most vulnerable. The federal government built and subsidized the infrastructure. The federal government must now address the fact that the infrastructure it built is generating legally defective financial instruments at industrial scale.

The Call to Action

This paper is addressed to every policymaker, regulator, legislator, financial stability official, and academic who has a role in protecting the integrity of the financial system. The risk identified here is not speculative. The statutes are enacted. The noncompliance is observable. The leverage is documented. The transmission mechanism is mapped. The only missing element is regulatory awareness—and this paper, together with SEC TCR #17729-304-613-692 and SEC TCR #17729-613-590-548, provides it.

FSOC must immediately assess medical debt as a financial stability risk. The $220 billion medical debt asset class has never been examined through a systemic risk lens (KFF & Peterson Health System Tracker, 2023). FSOC’s own Analytic Framework—assessing leverage, liquidity risk, interconnections, operational risk, and complexity/opacity—applies directly to the phantom capital transmission chain. The assessment must include the interaction between state consumer protection statutes and securitization structures.

The Federal Reserve must incorporate retroactive statutory nullification scenarios into stress testing. No current CCAR/DFAST scenario models the possibility that a class of consumer debt is rendered void by operation of statute. This is not an exotic tail risk. SB 1061 is enacted law. Fifteen to sixteen states are pursuing similar legislation (National Consumer Law Center, 2025). The stress testing framework must be updated to include a variable for retroactive statutory nullification of receivables within leveraged loan and CLO collateral pools.

Rating agencies must examine billing template compliance as a precondition of medical receivables analysis. No Moody’s, S&P, Fitch, DBRS, or KBRA methodology asks whether the billing communications generating healthcare company revenue comply with Regulation F and state disclosure requirements. This is the structural equivalent of rating mortgage-backed securities without examining whether the mortgages were properly originated. The rating agencies must develop methodology to assess statutory compliance risk at the receivable level—or publicly acknowledge that their current models do not capture this category of risk.

The OCC, FDIC, and state banking regulators must assess whether banks holding or lending against medical debt receivables face safety and soundness exposure from phantom capital contamination. Multiple banks have financed medical debt portfolios, provided revolving credit facilities to healthcare companies, and lent against medical receivables as collateral. If the phantom debt exposure is as widespread as the template-level noncompliance suggests, multiple institutions may simultaneously face collateral impairment, loss provisioning requirements, and capital adequacy pressure that no current supervisory examination is designed to detect.

Congress must hold hearings on the intersection of consumer protection enforcement and financial stability. The existing legislative framework treats consumer protection and financial stability as separate domains, administered by separate agencies, with separate analytical tools. Phantom capital demonstrates that they are the same domain—that a consumer protection statute can render a financial asset class void, and that the financial system has no mechanism for detecting or responding to this outcome. The legislative gap is not merely regulatory. It is architectural.

HHS, ONC, and CMS must immediately assess whether federally certified EHR platforms generate billing communications that comply with the FDCPA, Regulation F, and state consumer protection statutes. The HITECH Act authorized $36 billion in federal incentives to drive EHR adoption. ONC certified the platforms. CMS administered the payments. The revenue cycle management modules embedded in these certified platforms are generating noncompliant billing communications that create phantom debt at industrial scale. The federal government built and subsidized this infrastructure. It has a direct responsibility to audit whether the billing templates in the platforms it certified comply with the consumer protection statutes that govern every patient financial communication those platforms generate. If they do not—and the evidence in this paper indicates that they do not—then the federal EHR certification program has a foundational compliance defect that must be remediated before it generates further contamination of the financial system and further deprivation of patient rights.

12. The only structural intervention: nationalized health payment infrastructure

Section 11 identified the crisis. This section identifies the only structural intervention capable of preventing total collapse of the healthcare transaction infrastructure in the United States.

The regulatory actions called for in Section 11—FSOC assessment, stress test incorporation, rating agency reform, OCC and FDIC examination, Congressional hearings—are necessary and urgent. But they are diagnostic. They identify the contamination, quantify the exposure, and prepare the system for the losses that are coming. They do not solve the underlying problem. The underlying problem is that the revenue cycle management industry—the private, for-profit billing infrastructure that processes substantially all healthcare transactions in the United States—is architecturally incapable of generating compliant billing communications under the current statutory framework, and every attempt to remediate that noncompliance generates a knowledge event confirming prior violations across the platform’s entire provider network, accelerating the liability inversion rather than resolving it.

The Remediation Paradox

Consider what happens when an RCM platform attempts to cure the noncompliance. The platform updates its billing templates to include the mandatory Regulation F disclosures and SB 1061 language. This update is an implicit admission that prior templates did not contain this language. Every provider on the platform now has documented evidence that every billing communication generated before the update was noncompliant. Every patient who received a noncompliant communication has a statutory claim. The platform’s compliance team has just created the evidentiary foundation for a class action encompassing every patient billed through the platform in every jurisdiction where the prior template was used. The remediation does not cure the liability—it crystallizes it. The act of fixing the problem proves the problem existed.

This is not a recoverable situation for the RCM industry as currently structured. athenahealth’s 170,000+ providers processing 315 million claims per year, Epic’s dominant hospital market share, Cerner/Oracle Health, eClinicalWorks, Veradigm—each platform faces the same paradox. Remediate and confirm the liability, or continue noncompliant operations and compound it. There is no third option within the existing structure. No amount of private capital will recapitalize these platforms into the same liability exposure. No acquirer will purchase an RCM company whose entire historical billing output is a generator of statutory damages. The industry’s equity value trends toward zero as awareness of the contamination spreads, and no market mechanism reverses that trajectory because the liabilities are statutory and cumulative—they grow with every billing cycle that has already occurred.

The Infrastructure Failure Scenario

When the RCM platforms become insolvent—not if, but when, because cumulative inverse liabilities are unbounded and grow with every historical billing cycle—the transaction infrastructure for healthcare billing in the United States ceases to function. Providers cannot generate bills. Insurance claims cannot be processed. Patient accounts cannot be managed. Revenue stops flowing to hospitals, surgery centers, physician practices, and post-acute care facilities. Payroll cannot be met. Supply chains break. And patients do not stop needing medical care because the billing software company went bankrupt.

This is not a financial crisis in the traditional sense. It is an operational crisis—a failure of the physical infrastructure that connects healthcare delivery to healthcare payment. The 2008 crisis froze credit markets but the underlying economy continued to function, however impaired. The phantom capital crisis, taken to its logical conclusion, breaks the mechanism by which healthcare providers are paid for delivering care. The question is not whether investors lose money. Investors are already exposed to losses that no model has priced. The question is whether Americans continue to have access to functioning healthcare when the billing layer between patients, providers, and insurers collapses.

Why No Private-Sector Solution Exists

A private-sector replacement for the RCM industry would need to build a compliant billing platform from scratch, assume none of the predecessor platforms’ liabilities, achieve nationwide provider adoption, and do so under a statutory framework that renders any noncompliant communication void and any remediation effort self-incriminating. The capital requirements are enormous, the liability exposure from any error is existential, and the time horizon is incompatible with the speed at which the current infrastructure is deteriorating. No venture capital firm, no private equity fund, and no strategic acquirer will underwrite this risk. The private market has priced this sector at uninvestable.

The insurance industry cannot solve this independently. Insurers process claims through the same RCM infrastructure. The hospital systems cannot solve this independently—they are the providers whose balance sheets are contaminated with phantom receivables and whose vendor relationships are with the same insolvent platforms. The problem is structural, industry-wide, and beyond the capacity of any private actor or consortium of private actors to resolve within the timeframe required to prevent the infrastructure failure scenario described above.

The Single-Payer Solution: TrumpCare

There is exactly one intervention that eliminates the contamination vector, replaces the failing infrastructure, and ensures continuity of healthcare access for all Americans: nationalization of health payment processing infrastructure under a single-payer framework.

A single-payer system—call it TrumpCare, call it Medicare for All, call it whatever name any administration wants to put on it—eliminates the patient-facing billing layer entirely. No patient-facing debt contracts means no Regulation F obligations. No Regulation F obligations means no SB 1061 void-contract triggers. No void-contract triggers means no phantom debt, no phantom receivables, no phantom capital, no liability inversion, no leveraged contingent liability inversion. The contamination vector ceases to exist because the category of transaction that generates it ceases to exist. The government becomes the single payer to healthcare providers, replacing the failed private billing infrastructure with a direct payment channel that does not route through patient debt contracts and therefore cannot generate the statutory liabilities that are destroying the current system.

The author of this paper would be happy for any policy from any administration that solves this problem before all Americans lose access to healthcare. This is not a political statement. It is the logical conclusion of the facts presented in Sections 1 through 11. The private RCM industry is collapsing under cumulative inverse liabilities. No private capital will replace it. No regulatory action can cure retroactive voidness. The only remaining option is a public payment infrastructure that bypasses the patient debt layer entirely. If President Trump wants to sign it, he should sign it. Let him take the credit. The author does not care whose name is on the legislation. The author cares that people do not die because the billing software company went bankrupt.

Conclusion: A Gap Confirmed Across Every Domain

This literature review confirms, through exhaustive search across ten domains and supplementary investigation of whistleblower program history and EHR infrastructure, that no prior academic work, regulatory framework, rating agency methodology, or financial stability analysis has identified or modeled the specific risk described in Phantom Capital. The gap sits at the intersection of well-established fields—contract law (void ab initio doctrine, centuries old), consumer protection (SB 1061, enacted September 24, 2024), structured finance (CLOs, $1.3–1.4 trillion market; LSTA, 2024), information economics (Akerlof, 1970; Gorton & Holmström, 2020), leverage propagation (Brunnermeier & Pedersen, 2009; Geanakoplos, 2010), systemic risk measurement (Acharya et al., 2017), and federal health information technology policy (HITECH Act, 2009)—each highly developed in isolation but never connected through the mechanism of retroactive statutory nullification.

This paper introduces nine novel terms—none of which existed in any prior academic, regulatory, legal, or financial publication—and identifies a systemic risk that is not theoretical, not emerging, and not contingent on future events. It is present in the financial system now. What follows is a summary of every dimension of the gap this paper has confirmed.

The contamination is unknowable, not merely unknown. The information needed to sort phantom receivables from performing receivables was never generated at origination. This is structural information absence—a condition no existing information economics framework has described, because every prior framework (Akerlof, 1970; Stiglitz & Weiss, 1981; Keys et al., 2010; Gorton & Holmström, 2020) assumes the sorting information exists somewhere. This is not a null value in an existing schema. It is the absence of the schema itself.

The recovery rate is not low. It is zero—and then it goes negative. Void contracts create no enforceable claim, unlike defaulted contracts that yield 60–70% recovery (Moody’s Investors Service, 2023). But zero is not the floor. Phantom debt is an inverse obligation: it generates statutory liabilities with every collection communication, every payment extracted, every credit report furnished. The loss-given-default exceeds 100% because the “asset” is a net liability. No financial model in any domain has a variable for an asset that becomes a liability through routine operation.

The contamination is correlated at the platform level, not the provider level. The diversification benefits that CLO structures depend upon—the assumption that medical receivables from different providers in different states represent independent credit risks—collapse because the receivables were generated by the same noncompliant billing templates on the same RCM platforms (athenahealth, Inc., 2024; Coval et al., 2009). The federal EHR adoption mandate (HITECH Act, 2009) created this correlation by consolidating American healthcare billing onto a handful of dominant platforms. The consolidation the government promoted as an efficiency gain is the mechanism that makes the crisis systemic.

No stress test, capital model, or systemic risk framework contains a variable for retroactive statutory nullification. Federal Reserve stress tests (CCAR/DFAST) model macroeconomic scenarios. Basel III/IV models credit risk, market risk, and operational risk. FSOC’s Analytic Framework assesses leverage, liquidity, interconnections, and complexity. None includes a legislative event that renders a class of debt instruments legally void. The risk is not merely unpriced—it is uncategorized.

No rating agency has examined billing template compliance as a precondition of medical receivables analysis. Moody’s, S&P, Fitch, DBRS, KBRA, and Scope all model structural isolation in securitization without ever asking whether the underlying receivable is legally valid (Fender & Kiff, 2004; Hill, 2010; Rösch & Scheule, 2011). No rating agency has published any commentary on SB 1061. The legal analysis flows downward from structure to transaction documents but never flows upward from the underlying obligation to question its legal existence.

No intervention plan exists for an asset class crisis where the asset does not exist. In 2008, the government bought the distressed assets. In 2020, the government injected liquidity. In the phantom capital crisis, there is nothing to buy, nothing to inject, and nothing to freeze—because the contracts that would create the obligations were never legally formed (Brunnermeier, 2009). The entire toolkit of crisis intervention—TARP, quantitative easing, guarantees, forbearance, restructuring—presupposes a real instrument with recoverable value. No central bank can buy what does not exist.

The contagion mechanism resembles a pandemic more than a traditional financial crisis. The contamination is already present in the system. It spreads through discovery—from awareness to awareness—as each analyst, rating agency, and CLO manager who examines a billing template confirms that the contamination exists everywhere the same noncompliant templates were used. The act of testing does not create the infection. It reveals an infection that has been spreading undetected.

Private enforcement of consumer protection statutes can trigger systemic financial consequences, and no scholarship has examined this intersection. Berger et al. (2022) noted that even public enforcement’s relationship to systemic risk was uninvestigated prior to their paper. SB 1061’s void-contract provision is enforceable through private litigation. The Georgia GFLA episode demonstrates that the financial system is acutely sensitive to private enforcement vectors—yet scholarship treated it as a lobbying story, not a financial stability event.

The federal EHR infrastructure—built and subsidized with $36 billion in HITECH Act incentives—is the transmission vector. ONC certified the clinical modules of these platforms. The billing modules—the modules generating phantom debt—were not subject to the same certification rigor. The federal government certified the clinical side and ignored the financial side, and the financial side is the one contaminating the global financial system.

The medical debt asset class was built on the systematic exploitation of vulnerable patients. Medical debt is the industrialized repackaging of human suffering into a tradeable financial instrument, generated at the moment a human being at their lowest and most vulnerable is presented with a billing document they did not negotiate and could not have negotiated. The omission of mandatory disclosure language is not an oversight. It is an architecture designed to maximize extraction and suppress the assertion of patient rights. SB 1061 and 15–16 similar state laws are the legislative recognition that this system must be dismantled (National Consumer Law Center, 2025).

The distributed compliance audit eliminates the enforcement bottleneck that protected noncompliant billing practices. Frontier AI enables any patient to upload a billing statement and receive a complete statutory analysis with per-violation damages calculations and enforcement-ready regulatory filings. The traditional assumption that only a small fraction of statutory violations will ever be enforced—the assumption on which the economic viability of noncompliant billing practices depends—is invalidated. The barrier to enforcement drops from “retain a consumer protection attorney” to “upload a PDF.” This methodology is published, replicable, and cannot be suppressed (Ibiza, 2026a, 2026b).

The author of this paper has no financial interest in any entity discussed herein, has twice requested SEC trading surveillance on himself—an act without precedent in the history of U.S. securities whistleblowing (Platt, 2024)—and swears under penalty of perjury that every statement in this paper and its accompanying disclosure is true and correct.

The $220 billion in outstanding US medical debt represents the substrate. California’s SB 1061, with 15–16 states following similar paths, represents the catalyst. The federal EHR infrastructure that the government built and subsidized represents the correlation mechanism. And the complete absence of any model, framework, intervention plan, or analysis capable of detecting, pricing, or responding to this risk represents the vulnerability that Phantom Capital is the first work to identify.

Andrex Ibiza | Waukesha, Wisconsin | March 11, 2026

Discoverer of Retroactive Statutory Nullification, Phantom Debt, Inverse Obligation, Phantom Receivables, Phantom Capital, Liability Inversion, Leveraged Contingent Liability Inversion, Structural Information Absence, and the Distributed Compliance Audit

Finance MBA — Johnson & Wales University

Independent Financial Analyst

Disclosure

I, Andrex Ibiza, the sole author of this paper, declare under penalty of perjury under the laws of the United States of America and the State of Wisconsin that all statements in this disclosure are true and correct to the best of my knowledge, information, and belief. I make this declaration voluntarily, without reservation, and with full awareness that any material misstatement herein may expose me to criminal liability under 18 U.S.C. § 1001 and applicable state law. I have previously submitted two sworn SEC whistleblower filings—TCR #17729-304-613-692 and TCR #17729-613-590-548—each certified under penalty of perjury pursuant to 28 U.S.C. § 1746. Those filings are irretractable federal records. This disclosure carries the same weight.

No Financial Interest. I hold no position—long, short, neutral, direct, indirect, beneficial, or constructive—in any equity, debt, derivative, option, warrant, swap, futures contract, forward contract, convertible instrument, or other financial instrument of any company, entity, or security referenced in this paper. I hold no position in any index, exchange-traded fund, mutual fund, or other pooled investment vehicle that is concentrated in or specifically tracks any entity discussed herein. This declaration covers, without limitation: Surgery Partners, Inc. (NASDAQ: SGRY); Commerce Bancshares, Inc. (NASDAQ: CBSH); UnitedHealth Group Incorporated (NYSE: UNH); Optum, Inc. and all UnitedHealth Group (NYSE: UNH) subsidiaries; Bain Capital, LP and all affiliated funds, general partners, portfolio companies, and co-investment vehicles; Hellman & Friedman LLC and all affiliated funds and portfolio companies; athenahealth, Inc.; Epic Systems Corporation; Oracle Health (formerly Cerner Corporation, a subsidiary of Oracle Corporation [NYSE: ORCL]); eClinicalWorks, LLC; Veradigm Inc. (formerly Allscripts Healthcare Solutions); and any collateralized loan obligation, asset-backed security, credit-linked note, total return swap, credit default swap, exchange-traded fund, or other structured product that may hold, reference, or derive value from the debt or equity of any entity discussed herein.

No Derivative Exposure. I hold no derivative position of any kind—including but not limited to put options, call options, warrants, rights, credit default swaps, total return swaps, equity swaps, variance swaps, contracts for difference, binary options, or any synthetic, exotic, or over-the-counter instrument—referencing any entity, security, index, or asset class discussed in this paper. I am party to no agreement, arrangement, understanding, or scheme that would provide me with economic exposure, whether positive or negative, to the market price, credit quality, or financial condition of any entity discussed herein.

No Short Selling. I have not sold short, nor entered into any arrangement, agreement, or understanding—whether directly, through an agent, through a nominee, or through any other intermediary—that would profit from a decline in the market value, credit rating, share price, bond price, or financial condition of any entity discussed in this paper. I have no pending order, standing instruction, limit order, stop order, contingent order, or algorithmic instruction to establish any such position. I have not lent, borrowed, or arranged to borrow any securities of any entity discussed herein.

No Compensation From Interested Parties. I have received no compensation, consideration, payment, fee, grant, gift, loan, equity stake, revenue share, referral fee, consulting fee, retainer, or anything of value—whether monetary or non-monetary, direct or indirect, current or deferred—from any party with a financial interest, competitive interest, regulatory interest, or litigation interest in any entity, security, or asset class discussed in this paper. No party has funded, sponsored, commissioned, directed, or otherwise supported or influenced this research. My sole financial expenditure in connection with this work is a personal subscription to Anthropic’s Claude AI platform, paid from my own funds.

No Advisory or Intermediary Relationship. I am not acting as an investment adviser, broker-dealer, financial intermediary, placement agent, solicitor, or finder. This paper does not constitute investment advice, a recommendation to buy, sell, hold, or avoid any security, or a solicitation of any transaction. I am not registered with the Securities and Exchange Commission, the Financial Industry Regulatory Authority (FINRA), the Commodity Futures Trading Commission (CFTC), the National Futures Association (NFA), or any state securities or commodities regulator in any advisory, dealer, or intermediary capacity. I am not affiliated with any registered investment adviser, broker-dealer, bank, insurance company, or other financial institution.

Pending Regulatory Filings. I have filed three whistleblower complaints with the Securities and Exchange Commission under the Dodd-Frank Wall Street Reform and Consumer Protection Act: SEC TCR #17729-304-613-692 (concerning Commerce Bancshares, Inc. [NASDAQ: CBSH], UnitedHealth Group Incorporated [NYSE: UNH], ProHealth Care, and Optum, Inc.) and SEC TCR #17729-613-590-548 (concerning Bain Capital, LP, athenahealth, Inc., and Surgery Partners, Inc. [NASDAQ: SGRY]). My 3rd TCR includes the full text of this article, TCR #17732-878-226-191.

A TCR—Tip, Complaint, or Referral—is the SEC’s formal intake instrument for whistleblower submissions. It is submitted under penalty of perjury: the filer certifies under 18 U.S.C. § 1001 and 28 U.S.C. § 1746 that the information provided is true and correct, and a materially false statement in a TCR is a federal crime punishable by up to five years’ imprisonment. Both of my TCRs were submitted with this certification. They are irretractable federal records in the possession of the Commission.

I may be entitled to a whistleblower award under Section 21F of the Securities Exchange Act of 1934 (15 U.S.C. § 78u-6) if the Commission brings a successful enforcement action resulting in monetary sanctions exceeding $1,000,000 based in whole or in part on information provided in those filings. I disclose this potential future contingent financial interest for completeness. Any such award is entirely contingent on independent SEC enforcement action, is not within my control, and does not constitute a present financial interest in any security or entity.

Personal Claims as Patient. I have pending settlement demands and regulatory complaints against certain entities discussed in this paper, including but not limited to athenahealth, Inc. (a Bain Capital / Hellman & Friedman portfolio company) and entities within the UnitedHealth Group Incorporated (NYSE: UNH) corporate family, arising exclusively from my direct experience as a patient who received medical services from providers using these entities’ billing platforms and their affiliates. These claims are based on statutory violations in my own medical billing. They are personal consumer claims under the FDCPA, Regulation F, and applicable state consumer protection statutes. They are categorically distinct from any financial market position. My analysis of systemic financial risk in this paper is based on publicly available information, statutory text, regulatory filings, and academic literature, not on material nonpublic information obtained through any privileged or confidential relationship.

No Trading Intent. I have no present intention—and have formed no plan, arrangement, understanding, or instruction—to acquire, dispose of, or establish any position in any security, derivative, or financial instrument of any entity discussed in this paper. I have not communicated trading recommendations to any third party. Should my circumstances change materially, any future position or trading activity would be disclosed in subsequent publications and, to the extent required by law, in regulatory filings.

Voluntary Request for SEC Trading Surveillance. In connection with my three sworn SEC whistleblower filings—each submitted under penalty of perjury—(TCR #17729-304-613-692, TCR #17729-613-590-548, and TCR #17732-878-226-191), I have thrice requested that the Securities and Exchange Commission conduct trading surveillance on my own accounts and identity. I made these requests voluntarily and without obligation, for the sole purpose of providing the Commission with independent, verifiable confirmation that I hold no position in and have executed no trades involving any security of any entity referenced in my filings or in this paper. The following literature review confirms that this act is without documented precedent in the history of U.S. securities whistleblowing.

No Precedent Exists for Voluntary Whistleblower Self-Surveillance: A Literature Review

Across the full history of the SEC whistleblower program (2010–present)—more than 80,000 tips received and over 400 individual award determinations totaling more than $2 billion in awards (SEC Office of the Whistleblower, 2024)—no publicly documented case exists of any whistleblower requesting that the Commission conduct trading surveillance on the whistleblower’s own accounts to verify the absence of a financial position in a reported entity. The concept is absent from SEC regulations, enforcement actions, award orders, Office of Inspector General reports, academic literature, legal commentary, and news media. No analogous case has been identified in any other U.S. whistleblower program.

The SEC’s Regulatory Framework Does Not Contemplate the Concept. No provision in SEC Rules 21F-1 through 21F-18 creates a mechanism for whistleblowers to voluntarily submit to trading surveillance or offer their own brokerage records for audit. The Form TCR (Tip, Complaint, or Referral)—which every whistleblower must file under penalty of perjury, certifying the truthfulness of its contents under 18 U.S.C. § 1001—asks about the tipster’s relationship to the reported entity but contains no field for disclosing one’s own trading positions, certifying the absence of a financial interest, or requesting that the SEC monitor one’s accounts. The SEC’s award determination criteria under Rule 21F-6 do consider whether a whistleblower “financially benefited” from the reported violations, but this assessment is performed by the Commission after the fact—a retrospective evaluation, not a proactive offer of transparency from the whistleblower.

Prominent Whistleblowers Moved in the Opposite Direction. The most instructive cases involve whistleblowers who had clear financial entanglements with the entities they reported—and who either disclosed those interests or accepted the ambiguity rather than requesting surveillance. Harry Markopolos, the most famous SEC whistleblower for his decade-long campaign to expose Bernie Madoff’s Ponzi scheme, later published a 2019 report alleging accounting fraud at General Electric Company (NYSE: GE). He openly disclosed a financial relationship with a hedge fund holding a short position in GE stock. GE’s CEO accused him of market manipulation. At no point did Markopolos request that the SEC surveil his trading. Carson Block of Muddy Waters Research received a $14 million SEC whistleblower award for his 2011 tip on Focus Media, despite simultaneously holding short positions (Bloomberg Law, 2022). The SEC’s own Office of the Whistleblower staff reportedly recommended against the award, but the Commission overruled them. Block was simultaneously under DOJ investigation for potential stock manipulation. No self-surveillance request was made or reported. Hindenburg Research disclosed short positions (or their absence) in published reports—for example, explicitly stating it held no position in Clover Health (NASDAQ: CLOV) in 2021 (Hindenburg Research, 2021)—but voluntary public disclosure is a fundamentally different act from requesting regulatory surveillance of one’s own trading accounts. Eric Ben-Artzi, the Deutsche Bank AG (NYSE: DB) whistleblower who rejected his $8.25 million award in 2016 on principle (Financial Times, 2016), demonstrated extraordinary ethical commitment but did not request self-surveillance.

Academic Scholarship Identified the Gap but Not the Solution. The most relevant scholarly work is Platt’s (2024) article in the Washington Law Review, “The Shortseller Enrichment Commission? Whistleblowers, Activist Short Sellers, and the New Privatization of Public Enforcement.” Platt documented that approximately 40% of SEC whistleblower bounty awards in recent years have gone to “outsiders,” including activist short sellers. He identified a critical structural gap: current SEC program rules do not require any accounting of tipsters’ private trading profits when making awards. Platt recommended the SEC amend its rules to require such an accounting—but this proposed reform operates in the opposite direction from self-surveillance. It envisions the SEC imposing a requirement on whistleblowers, not whistleblowers proactively volunteering transparency. The fact that leading scholarship frames the problem as a regulatory gap rather than something any whistleblower has addressed on their own initiative underscores the novelty of the concept.

No Parallel Exists in Any Other U.S. Whistleblower Program. The research extended beyond the SEC to every major U.S. whistleblower framework. The CFTC whistleblower program mirrors the SEC’s structure and similarly lacks any mechanism for self-surveillance requests. The IRS Whistleblower Office (Form 211) requires disclosure of the whistleblower’s relationship to the subject but not proof of lack of financial conflict. False Claims Act (qui tam) relators must deliver substantially all material evidence of the defendant’s fraud, but this concerns evidence of wrongdoing, not the relator’s own financial posture. No case was found of a qui tam relator proactively submitting personal financial records to prove absence of conflict. No documented case of a FINRA whistleblower requesting self-surveillance was identified. Across all programs, no formal mechanism exists for a whistleblower to proactively prove they have no financial stake by submitting to regulatory surveillance.

The absence of any precedent is striking given the structural incentives at play. The tension between financial incentives and whistleblower credibility is one of the most heavily debated issues in securities enforcement scholarship. Short sellers who double as whistleblowers have drawn scrutiny from Congress, the DOJ, the SEC itself, and foreign regulators. Whistleblower attorneys, who play a central role in tip preparation, have no incentive to encourage clients to open their financial lives to regulatory scrutiny beyond what is required. The SEC has not created incentive structures that reward such transparency—proactive disclosure of a clean trading record does not formally enhance an award under Rule 21F-6. The concept of a whistleblower voluntarily requesting regulatory trading surveillance on themselves was without documented precedent across the entire landscape of U.S. whistleblower programs—until now. The author of this paper has made the request thrice, under oath, in connection with three sworn SEC whistleblower filings (TCR #17729-304-613-692, TCR #17729-613-590-548, and TCR #17732-878-226-191), and is now swearing it a third time in the declaration that follows. The existing regulatory infrastructure is not designed to process this act of transparency.

I did it anyway.

References

Acharya, V. V., Pedersen, L. H., Philippon, T., & Richardson, M. (2017). Measuring systemic risk. Review of Financial Studies, 30(1), 2–47.

Acharya, V. V., Schnabl, P., & Suarez, G. (2013). Securitization without risk transfer. Journal of Financial Economics, 104(3), 515–536.

Adrian, T., & Ashcraft, A. B. (2010). Shadow banking: A review of the literature (Staff Report No. 458). Federal Reserve Bank of New York.

Adrian, T., & Ashcraft, A. B. (2012). Shadow banking regulation (Staff Report No. 559). Federal Reserve Bank of New York.

Akerlof, G. A. (1970). The market for “lemons”: Quality uncertainty and the market mechanism. Quarterly Journal of Economics, 84(3), 488–500.

Alyeska Pipeline Service Co. v. Wilderness Society, 421 U.S. 240 (1975).

American Medical Association, Council on Medical Service. (2023). Report 7: Addressing medical debt.

Arner, D. W., Barberis, J., & Buckley, R. P. (2017). FinTech, RegTech, and the reconceptualization of financial regulation. Northwestern Journal of International Law & Business, 37(3), 371–413.

Ashcraft, A. B., & Schuermann, T. (2008). Understanding the securitization of subprime mortgage credit (Staff Report No. 318). Federal Reserve Bank of New York.

athenahealth, Inc. (2024). Company overview and platform statistics.

https://www.athenahealth.com

Bain Capital. (2022, February 15). Bain Capital and Hellman & Friedman complete acquisition of athenahealth [Press release].

Bain Capital. (2024). Portfolio: Surgery Partners.

https://www.baincapital.com

Benmelech, E., & Dlugosz, J. (2009). The alchemy of CDO credit ratings. Journal of Monetary Economics, 56(5), 617–634.

Berger, A. N., Cai, J., Roman, R. A., & Sedunov, J. (2022). Supervisory enforcement actions against banks and systemic risk. Journal of Banking & Finance, 140, 106222.

Bloomberg. (2024). CLO ETF assets surpass $19 billion. Bloomberg News.

Bloomberg Law. (2022, July 29). Mystery of Carson Block SEC payout deepens with suit outing him. Bloomberg Law.

Bowen v. Georgetown University Hospital, 488 U.S. 204 (1988).

Board of Governors of the Federal Reserve System. (2025, November). Financial stability report.

Breeden, S. (2022, November). Risks from leverage: How did a small corner of the pensions industry threaten financial stability? [Speech]. Bank of England.

Brunnermeier, M. K. (2009). Deciphering the liquidity and credit crunch 2007–2008. Journal of Economic Perspectives, 23(1), 77–100.

Brunnermeier, M. K., & Pedersen, L. H. (2009). Market liquidity and funding liquidity. Review of Financial Studies, 22(6), 2201–2238.

Carlino, T., Foley-Fisher, N., Heinrich, N., & Verani, S. (2025, March). Life insurers and CLO regulatory capital arbitrage (FEDS Notes). Board of Governors of the Federal Reserve System.

Consumer Data Industry Association v. Frey, 26 F.4th 1 (1st Cir. 2022).

Consumer Financial Protection Bureau. (2015). CFPB sues Universal Debt & Payment Solutions for phantom debt collection [Press release].

Consumer Financial Protection Bureau. (2022). Medical debt burden in the United States.

Cooper v. Aaron, 358 U.S. 1 (1958).

Cordell, L., Roberts, M. R., & Schwert, M. (forthcoming). CLO performance. Journal of Finance. (Philadelphia Fed Working Paper 20-48).

Coval, J. D., Jurek, J. W., & Stafford, E. (2009). The economics of structured finance. Journal of Economic Perspectives, 23(1), 3–25.

Damgaard, J., Elkjaer, T., & Johannesen, N. (2019). What is real and what is not in the global FDI network? IMF Working Paper, WP/19/274.

Crémer, J., & Khalil, F. (1992). Gathering information before signing a contract. American Economic Review, 82(3), 566–578.

Dang, T. V., Gorton, G., & Holmström, B. (2020). The information view of financial crises. Annual Review of Financial Economics, 12, 39–65.

Durand, C. (2017). Fictitious capital: How finance is appropriating our future. Verso.

Esty, B. C. (1998). The impact of contingent liability on commercial bank risk taking. Journal of Financial Economics, 47(2), 189–218.

Federal Trade Commission. (2019). FTC-CFPB Operation Corrupt Collector [Press release].

Fender, I., & Kiff, J. (2004). CDO rating methodology: Some thoughts on model risk and its implications. BIS Working Papers, No. 163.

Financial Stability Board. (2023). The financial stability implications of leverage in non-bank financial intermediation.

Financial Times. (2016, August 16). Deutsche Bank whistleblower refuses $8.25m SEC award. Financial Times.

Foley-Fisher, N., Gorton, G., & Verani, S. (2020). Adverse selection dynamics in privately-produced safe debt markets (NBER Working Paper No. 28016).

Fulford, S. L., & Wilson, C. (2025). Medical debt in America. Health Affairs Scholar.

Geanakoplos, J. (2010). The leverage cycle. NBER Macroeconomics Annual, 24(1), 1–65.

Gerding, E. F. (2009). The subprime crisis and the link between consumer financial protection and systemic risk. Florida International University Law Review, 4(2).

Gorton, G. B. (2010). Slapped by the invisible hand: The panic of 2007. Oxford University Press.

Gorton, G., & Holmström, B. (2015). Information-insensitive debt. Unpublished manuscript.

Griffin, J. M., & Nickerson, J. (2023). Are CLO collateral and tranche ratings disconnected? Review of Financial Studies, 36(6), 2319+.

Grossman, R. S. (2013). Contingent liability and risk taking in British banking, 1850–1913. Economic History Review, 66(1), 271–291.

Hart, O., & Moore, J. (1990). Property rights and the nature of the firm. Journal of Political Economy, 98(6), 1119–1158.

Hilferding, R. (1910). Finance capital. (T. Bottomore, Trans., 1981). Routledge & Kegan Paul.

Hill, C. A. (2010). Why did rating agencies do such a bad job rating subprime securities? University of Pittsburgh Law Review, 71, 585–608.

Hindenburg Research. (2021, February 4). Clover Health: How the “Medicare Disruptor” is ensnaring patients.

Ibiza, A. (2026a, March 8). The last line of defense. Flat Crotch Dispatch.

Ibiza, A. (2026b, March 11). The distributed compliance audit. Flat Crotch Dispatch.

Kaiser Family Foundation & Peterson Health System Tracker. (2023). Health costs: Americans’ challenges with health care costs.

Keys, B. J., Mukherjee, T., Seru, A., & Vig, V. (2010). Did securitization lead to lax screening? Quarterly Journal of Economics, 125(1), 307–362.

Knight, F. H. (1921). Risk, uncertainty, and profit. Houghton Mifflin.

Landgraf v. USI Film Products, 511 U.S. 244 (1994).

LCD/Pitchbook. (2024). Leveraged lending annual review.

Levitin, A. J., & Wachter, S. M. (2012). Explaining the housing bubble. Georgetown Law Journal, 100(4), 1177–1258.

Levitin, A. J., & Wachter, S. M. (2020). The great American housing bubble. Harvard University Press.

Loan Syndications and Trading Association. (2024). CLO market review.

Marx, K. (1894). Capital: Volume III. (F. Engels, Ed.).

Mian, A., & Sufi, A. (2009). The consequences of mortgage credit expansion. Quarterly Journal of Economics, 124(4), 1449–1496.

Moody’s Investors Service. (2022). athenahealth, Inc.: Credit opinion.

Moody’s Investors Service. (2023). Annual default study: Corporate default and recovery rates, 1920–2023.

National Consumer Law Center. (2025). State medical debt protection legislation tracker.

Newman v. Piggie Park Enterprises, Inc., 390 U.S. 400 (1968).

Office of Financial Research. (2024). Financial Research Advisory Committee: CLO market data.

Oei, S.-Y., & Ring, D. M. (2018). The tax lives of Uber drivers: Evidence from internet discussion forums. Columbia Journal of Tax Law, 8(1), 56–112.

Platt, A. I. (2024). The shortseller enrichment commission? Whistleblowers, activist short sellers, and the new privatization of public enforcement. Washington Law Review.

Quinlan, T., & Grein, S. S. (2023, December 4). BNPL: The phantom debt. Wells Fargo Economics.

Rajan, U., Seru, A., & Vig, V. (2015). The failure of models that predict failure. Journal of Financial Economics, 115(2), 237–260.

Rösch, D., & Scheule, H. (2011). Securitization rating performance and agency incentives. BIS Papers, No. 58.

Scheuerman, S. S. (2011). Emerging issues in statutory damages. Jones Day.

Securities and Exchange Commission, Office of the Whistleblower. (2024). Annual report to Congress on the Dodd-Frank whistleblower program.

Sobol, R. B. (2014). Protecting consumers from zombie-debt collectors. New Mexico Law Review, 44, 327–370.

Stigler, G. J. (1961). The economics of information. Journal of Political Economy, 69(3), 213–225.

Stiglitz, J. E., & Weiss, A. (1981). Credit rationing in markets with imperfect information. American Economic Review, 71(3), 393–410.

Surgery Partners, Inc. (2025). Annual report on Form 10-K for the fiscal year ended December 31, 2024. SEC EDGAR.

ZAIS Group. (2024). Healthcare sector concentration in leveraged loan and CLO markets [Research note].

The Controlled Event

A compliance audit, in the traditional corporate sense, is a controlled event. It has a scope. It has a timeline. It has a budget. It has a counterparty you can negotiate with. The auditors arrive on a date you agreed to. They examine the documents you prepared for them. They produce findings you can respond to before the report is finalized. The entire apparatus of corporate compliance—from the Big Four accounting firms to the in-house legal teams that manage them—is built around the assumption that audits are institutional, bounded, and manageable.

This assumption has been correct for as long as compliance analysis required institutional resources. An audit of billing templates across a healthcare system requires attorneys who know the relevant statutes, analysts who can review the documents, and enough time and budget to examine a representative sample. The cost of performing this work has always functioned as a natural barrier—not to compliance itself, but to the discovery of noncompliance. When the cost of checking whether a billing statement contains every legally required disclosure is measured in thousands of dollars per hour of attorney time, the probability that any individual patient will check is effectively zero.

Corporate risk models depend on this probability. The expected cost of noncompliance is a function of the probability of detection multiplied by the cost of the violation if detected. When the probability of detection is near zero, even catastrophic per-violation penalties produce a manageable expected cost. This is not a secret. It is the foundational logic of every cost-benefit analysis that has ever concluded it was cheaper to remain noncompliant than to update the templates. The math has always worked. The filtration rate between “statements sent” and “legally sophisticated complaints received” was so extreme that portfolio-wide noncompliance was an economically rational strategy.

The filtration rate between statements sent and legally sophisticated complaints received was so extreme that portfolio-wide noncompliance was an economically rational strategy.

The distributed compliance audit breaks the funnel. It breaks it permanently, irreversibly, and in a way that no corporate legal department has a playbook to address. This essay explains why.

What a Distributed Compliance Audit Is

In “The Last Line of Defense,” I described the mechanism by which a single person with a twenty-dollar AI subscription can identify systemic statutory noncompliance in medical billing infrastructure. I described how uploading one billing statement to an artificial intelligence tool and asking whether it contained all legally required disclosures produced an answer in seconds—an answer that would have cost thousands of dollars in attorney time five years ago. I described how that single discovery, replicated across nine entities, revealed portfolio-wide noncompliance with federal and state statutes governing medical debt collection.

At the end of the essay, I issued a call to action: find the nearest medical billing statement in your house, upload it to any AI tool you have access to, and ask whether it contains every disclosure required by law in your state. Then come back and tell me what you found.

People did. And they found exactly what I found. Noncompliant. Noncompliant. Noncompliant.

Each of those independent discoveries is a distributed compliance audit. Not because the discoverers coordinated. Not because they followed a protocol designed by a regulatory body. Not because they were acting under the authority of any institution. They are a distributed compliance audit because they are independent actors, performing the same analytical function, on the same category of document, using the same category of tool, arriving at the same category of finding—and each one generating an independent evidentiary record that the document they examined does not comply with the law that governs it.

The word “distributed” is precise. In computer science, a distributed system is one in which processing occurs across multiple independent nodes rather than at a single central location. The nodes do not need to communicate with each other. They do not need to be aware of each other. They simply execute the same function on their own inputs and produce outputs that are independently valid. A distributed compliance audit operates identically. Each node—each person with a billing statement and an AI tool—executes the same analytical function and produces a finding that stands on its own.

The properties that make it ungovernable

A traditional compliance audit has properties that make it manageable from the perspective of the audited institution. It is centralized: one firm, one team, one report. It is bounded: a defined scope, a defined timeline, a defined set of documents. It is negotiable: the audited entity participates in scoping, responds to preliminary findings, and can influence the final output. It is singular: one audit produces one set of findings that can be addressed as a unit.

A distributed compliance audit has none of these properties.

It is acephalous. There is no leader. There is no coordinating body. There is no one to negotiate with, serve with an injunction, acquire, discredit, or co-opt. The audit is not an organization. It is a methodology being independently executed by an unknowable number of people.

It is asynchronous. The audits are happening at different times, in different places, on different documents, at a pace no institution can monitor or predict. There is no start date and no end date. There is no “audit period” to prepare for or survive.

It is self-replicating. Every person who discovers a noncompliant template and shares their finding—publicly, privately, in a social media post, in a conversation with a friend, in a regulatory complaint—creates the potential for additional audits by additional people. The methodology does not need to be taught in a formal sense. It can be described in a single sentence: upload your billing statement to an AI and ask if it’s compliant.

It is confirmatory, not adversarial. This is the most important property and the one corporate legal departments will struggle with most. The distributed audit is not making an argument. It is not advancing a legal theory. It is not asserting a position that can be rebutted. It is checking a binary factual condition: does the document contain the required disclosure language, yes or no. The AI is not exercising judgment. It is reading the document. The finding is not an opinion. It is an observation. And because the underlying condition is binary—the disclosure is either present or it is not—the confirmation rate across independent auditors will approach one hundred percent.

The distributed audit is not making an argument. It is checking a binary factual condition. The confirmation rate across independent auditors will approach one hundred percent.

This Is an All Fifty States Problem

A misconception that will be comforting to some readers and devastating when corrected: the distributed compliance audit is not a California problem. California’s SB 1061—Civil Code § 1785.27—is the most dramatic single-state mechanism because it renders noncompliant medical debt contracts void at formation with no cure provision, transforming every dollar of that debt into what I have called phantom debt—debt that appears real on paper but is legally nonexistent. SB 1061 is the nuclear trigger. But focusing exclusively on SB 1061 misses the broader architecture of exposure, which operates at three independent levels: federal law that applies in every state, state consumer protection statutes that exist in every jurisdiction, and the interaction effects between them.

The foundational layer is federal. The Consumer Financial Protection Bureau’s Regulation F—12 C.F.R. § 1006.34—requires specific mandatory validation disclosures in every debt collection communication sent to every consumer in the United States. These are not optional. They are not best practices. They are regulatory requirements with the force of federal law. Every billing statement, every collection letter, every communication that constitutes an attempt to collect a debt must include: a notice of the right to dispute within thirty days, a statement of the consequences of non-dispute, a tear-off dispute form complying with the regulation’s format requirements, a statement that verification will be provided on request, and identification of the original creditor. A billing template that omits these disclosures is noncompliant with federal law in every state, for every patient, on every statement it generates. There is no California-specific element to this. A noncompliant template in Tennessee generates the same Regulation F violations as a noncompliant template in California.

The Fair Debt Collection Practices Act—15 U.S.C. §§ 1692 et seq.—provides the private enforcement mechanism for Regulation F and imposes its own independent prohibitions. Section 1692e prohibits false, deceptive, or misleading representations. Section 1692f prohibits unfair or unconscionable collection methods. Section 1692g establishes the validation framework. Each violation is independently actionable. Statutory damages of up to $1,000 per action are available regardless of actual harm. Attorney fees are mandatory for prevailing plaintiffs under § 1692k. These provisions apply in all fifty states.

When a person in Wisconsin uploads a billing statement to an AI tool and discovers that the statement omits every mandatory Regulation F disclosure, they have identified a federal violation. They do not need SB 1061. They do not need to be a California patient. They need only the FDCPA, which has protected every American consumer since 1977.

The Wisconsin Consumer Act

But federal law is only the first layer. Every state has its own consumer protection framework, and many of them are more aggressive than the federal floor. Wisconsin is a particularly instructive example—not because it is the most dramatic, but because it illustrates how state law multiplies the exposure that federal law creates.

Wisconsin’s Consumer Act, codified at Chapters 421 through 427 of the Wisconsin Statutes, is one of the most comprehensive state-level consumer credit and debt collection frameworks in the country. Chapter 427—the debt collection chapter—operates concurrently with the FDCPA but imposes independent obligations and provides independent remedies.

Section 427.104 of the Wisconsin Consumer Act enumerates a detailed list of prohibited debt collection practices. These include threatening action that cannot legally be taken, communicating with the consumer’s employer about the debt except as permitted by law, disclosing or threatening to disclose information adversely affecting the consumer’s reputation when the collector knows or should know the information is false, engaging in conduct that can reasonably be expected to threaten or harass the consumer or members of the consumer’s family, using obscene or threatening language, and—critically—making false representations or implications of the nature or legal status of a claim. A debt collector that represents phantom debt as a valid obligation, or that pursues collection on a debt it has not validated, is making a false representation of the legal status of the claim under Wisconsin law independent of any federal violation.

The remedies under the Wisconsin Consumer Act are substantial and independent of FDCPA remedies. Section 425.308 provides that a consumer may recover actual damages, a penalty not exceeding $1,000 for each violation, and the costs of the action including reasonable attorney fees. The Act also permits class actions. And unlike the FDCPA, which has a one-year statute of limitations, the Wisconsin Consumer Act provides a longer window for enforcement. Wisconsin courts have held that the Consumer Act and the FDCPA operate independently—violations of one do not require violations of the other, and remedies under both can be pursued simultaneously.

But Chapter 427 is not the only Wisconsin weapon. Wisconsin’s Deceptive Trade Practices Act—Section 100.18 of the Wisconsin Statutes—independently prohibits any assertion, representation, or statement of fact that is untrue, deceptive, or misleading in connection with the sale of merchandise or services. Medical billing is a service. Pursuing collection on unvalidated debt using noncompliant communications constitutes a deceptive representation in connection with that service. Section 100.18(11)(b)(2) provides for twice the amount of pecuniary loss, together with costs and reasonable attorney fees. The doubling provision is statutory. It is not discretionary. It applies on top of—not in lieu of—the remedies available under Chapter 427 and the FDCPA.

So a single noncompliant billing statement received by a Wisconsin consumer generates: FDCPA violations with statutory damages of up to $1,000 per action plus mandatory attorney fees; Wisconsin Consumer Act Chapter 427 violations with penalties of up to $1,000 per violation plus attorney fees; and Wisconsin DTPA § 100.18 violations with mandatory doubling of pecuniary loss plus attorney fees. Three independent statutory frameworks. Three independent sets of damages. Three independent fee-shifting provisions ensuring that enforcement is economically viable.

A single noncompliant billing statement received by a Wisconsin consumer generates violations under three independent statutory frameworks, each with its own damages provisions and each with mandatory attorney fee-shifting. Federal law is the floor, not the ceiling.

The multiplication across fifty states

Wisconsin is not exceptional. It is illustrative. Every state has a consumer protection statute—a “mini-FTC Act” or equivalent—that prohibits unfair or deceptive practices in consumer transactions. Many have specific debt collection statutes that operate concurrently with the FDCPA. California has the Rosenthal Fair Debt Collection Practices Act and SB 1061. New York has General Business Law § 349. Texas has the Deceptive Trade Practices–Consumer Protection Act with treble damages for knowing violations. Illinois has the Consumer Fraud and Deceptive Business Practices Act. Massachusetts has Chapter 93A with automatic treble damages. The specifics vary. The structural pattern does not: federal violations create the floor, state law multiplies the exposure above it, and fee-shifting provisions at both levels ensure that private enforcement is economically viable.

This means the distributed compliance audit is not identifying a California problem that happens to have implications elsewhere. It is identifying a universal federal noncompliance problem that generates compounding state-law exposure in every jurisdiction where a noncompliant statement was received. The patient in Waukesha, Wisconsin who discovers a Regulation F violation on their billing statement has a federal claim and at least two independent state-law claims. The patient in Austin, Texas has a federal claim and a DTPA claim with potential treble damages. The patient in Boston has a federal claim and a Chapter 93A claim with automatic trebling. Every state adds its own layer. The aggregate exposure across fifty states is not fifty times one state’s exposure—it is the sum of fifty independent liability frameworks, each with its own calculation, each with its own fee-shifting, and each making enforcement independently economically viable regardless of what happens in any other state.

The institution using a noncompliant billing template is not facing a California lawsuit. It is facing a liability matrix that spans every state in which it sent a noncompliant communication, under every statute that applies in each of those states, with independent enforcement by independent plaintiffs in independent jurisdictions. The distributed compliance audit reveals this matrix one node at a time—but the matrix existed the moment the template was deployed.

THE ALL FIFTY STATES EXPOSURE

Every Violation Is a Bounty

There is a feature of the private attorney general framework that the previous section implies but does not state with sufficient force. It is the single most important structural fact for anyone performing a distributed compliance audit, and it is the fact that makes the entire system work: every violation you find gives you an affirmative legal claim against the entity that committed it. Not a grievance. Not a complaint that disappears into an agency’s triage queue. A legal claim—with statutory damages, with fee-shifting, and with standing to sue in civil court if the entity refuses to pay.

This is the bounty. Congress built it deliberately. When Congress enacted the FDCPA, it did not just prohibit unfair debt collection practices. It created a private right of action—meaning any person who is the subject of a violation can bring a lawsuit in their own name—and it attached a financial incentive: statutory damages of up to $1,000 per action, plus actual damages, plus mandatory attorney fees for the prevailing plaintiff. When Wisconsin enacted the Consumer Act, it created the same structure: statutory penalties of up to $1,000 per violation, plus actual damages, plus attorney fees. When Wisconsin enacted § 100.18 of the DTPA, it added mandatory doubling of pecuniary loss, plus attorney fees. Each of these provisions is a bounty. Each one says: if you find a violation, you are entitled to money, and the entity that violated the law pays for your lawyer.

The fee-shifting provision is the engine that makes the entire system operational. Under the American Rule, each party pays its own attorney fees. This means that a consumer with a $1,000 statutory damages claim would spend more on legal representation than the claim is worth—making enforcement economically irrational. Fee-shifting inverts that calculus. When the statute provides that the prevailing plaintiff recovers attorney fees from the defendant, the consumer’s cost of enforcement drops to zero. The attorney takes the case knowing that if they win, the defendant pays their fees. The consumer’s only cost is their willingness to be the plaintiff.

This is not theoretical. Consumer protection attorneys take FDCPA and state UDAP cases on a fee-shifting basis every day. The attorney evaluates the claim. If the violation is clear and the evidence is strong, the attorney takes the case because the fee award—paid by the defendant—will exceed the cost of prosecution. The consumer does not pay a retainer. The consumer does not pay hourly rates. The consumer shares in the settlement or judgment. The attorney is compensated by the violator. Everyone is aligned except the defendant, who is paying for the privilege of having violated the law.

The binary claim is undefendable

What makes the distributed compliance audit’s bounties uniquely powerful is that the underlying claims are binary. The Regulation F disclosure is either on the document or it is not. The SB 1061 language is either present in the contract or it is not. The Wisconsin Consumer Act’s prohibition on false representations of the legal status of a claim is either violated or it is not. There is no interpretive ambiguity. There is no good-faith defense. There is no reasonable-minds-can-differ argument. The document speaks for itself.

In litigation, the most expensive and uncertain phase is the dispute over facts. Did the defendant actually do what the plaintiff says they did? Was the conduct intentional? Were the representations misleading to a reasonable consumer? These questions consume months of discovery, depositions, expert testimony, and motion practice. They are where defendants spend millions of dollars to create enough uncertainty to force a favorable settlement or win at trial.

A binary compliance claim eliminates this entire phase. The evidence is the defendant’s own document. The question is whether the document contains the required language. The answer is observable by anyone who can read. There is no discovery dispute because there is nothing to discover—the document is the evidence, the plaintiff already has it, and the defendant cannot dispute its contents because the defendant created it. The defendant’s only option is to argue about the legal significance of the undisputed fact, and in the case of Regulation F’s mandatory disclosure requirements, that argument is foreclosed by the regulation itself.

This means the bounty is not speculative. It is not a theoretical claim that might survive summary judgment if the facts break right. It is a claim built on undisputed evidence that the defendant produced and cannot contradict. A consumer attorney evaluating whether to take this case on a fee-shifting basis does not need to assess the probability of winning on the facts. The facts are not in dispute. The only question is the measure of damages—and the statutory minimums provide a floor that makes the case economically viable regardless.

A binary compliance claim eliminates the most expensive phase of litigation. The evidence is the defendant’s own document. The defendant cannot dispute its contents because the defendant created it.

Template-level findings scale by following the money

Here is where the bounty becomes something larger than an individual remedy. When you find a noncompliant billing statement, you have identified a violation affecting your own account. But if the noncompliance is at the template level—if the missing disclosure is absent from the template that generates statements for all accounts, not just yours—then your individual finding is evidence of a violation affecting every person who received a statement generated by that template.

At that point, the question is no longer what your individual claim is worth. The question is how big the entity is. How many patients does it serve? How many clinics does it operate? How many statements did the noncompliant template generate? The answers to these questions are public. They are in SEC filings, annual reports, press releases, and trade publications. A hospital system that serves 500,000 patients annually and uses a noncompliant billing template has generated 500,000 violations per year. A billing platform that serves 140,000 providers has generated noncompliant templates for 140,000 practices. The math is multiplication, and the inputs are the entity’s own publicly reported operating metrics.

This is what transforms a billing dispute into a systemic enforcement action. You do not need inside information to calculate the scope of exposure. You need one noncompliant statement—your own—to establish the template-level deficiency, and then you need the entity’s publicly available data to estimate the number of affected accounts. The demand letter writes itself: here is the violation, here is the evidence that it is template-level, here is the number of patients you serve, here is the statutory damages calculation, here is the settlement demand that represents a fraction of your total exposure. The entity’s own scale becomes the leverage.

For publicly traded companies, the scaling exercise is even more powerful. SEC filings disclose revenue, patient volume, facility count, and geographic footprint. The entity’s own 10-K tells you how to calculate its exposure. And the moment you calculate that exposure and communicate it to the entity’s officers, you have created a knowledge event that implicates their Sarbanes-Oxley certifications—because they are now personally aware of a material contingent liability that is not reflected in the financial statements they signed.

You do not need a lawyer to sue someone

There is no law in the United States that requires you to have an attorney to file a civil lawsuit. None. You can walk into any civil court in any state, file a complaint, and prosecute your own case. This is not a loophole. It is a constitutional right. Courts are public institutions. They are open to every citizen. The notion that you need a lawyer’s permission to access them is a cultural myth that serves the legal profession’s economic interests, not yours.

For a binary compliance claim, the cultural myth is not just wrong—it is backwards. Think about what a lawyer actually does in a typical civil case. They investigate facts that are in dispute. They depose witnesses. They argue over the meaning of ambiguous contract terms. They navigate discovery fights over what documents must be produced. They file motions contesting the other side’s interpretation of events. Every one of these functions presupposes that there is something to argue about.

In a binary compliance claim, there is nothing to argue about. The disclosure is on the document or it is not. You do not need an attorney to determine this. You can read. The judge can read. You walk into the courtroom with your billing statement in one hand and the statute in the other. You show the judge the statement. You show the judge the statute. You point to where the mandatory disclosure language is required to appear. You point to where it does not appear on your statement. You are done. The defendant cannot dispute the contents of a document it created. The defendant cannot argue that the disclosure is present when anyone with eyes can see that it is not. The claim is proved the moment the document is admitted into evidence.

In this situation, having a lawyer is a tax. A consumer protection attorney who takes your case on a fee-shifting basis will share in your settlement. A consumer protection attorney who takes your case on contingency will take a third of your recovery. In either arrangement, the attorney is being compensated for expertise you do not need, because there is no ambiguity to resolve, no facts to investigate, and no legal theory to construct. The theory is: the law requires this disclosure, my statement does not have it, here is the statement. That is the entire case. You do not need three years of law school to make it.

Small claims courts in most states have jurisdiction over claims up to $5,000 or $10,000. Their procedures are explicitly designed for people without attorneys. The filing fees are nominal—often under $100. The hearings are informal. The rules of evidence are relaxed. And the defendant—the hospital system, the billing platform, the debt collection company—must still appear. If the defendant is an out-of-state entity, it must retain local counsel in your jurisdiction to respond to your small claims action. A lawyer billing $300 to $500 per hour. In your state. On your schedule. To defend against a claim the defendant cannot win on the facts.

The cost of the defendant’s legal representation to fight your small claims action will exceed the cost of settling with you before the first hearing. The defendant’s attorney knows this. The defendant’s general counsel knows this. The rational institutional response to a binary claim filed by a pro se plaintiff in the plaintiff’s home jurisdiction is to settle, because the alternative is spending $5,000 in attorney fees to lose a case that was unwinnable the moment the noncompliant document was printed.

In a binary compliance claim, having a lawyer is a tax. You do not need three years of law school to hold up a document and point to where the required language is not.

When a lawyer makes sense

A lawyer becomes valuable—not necessary, but valuable—when the scale of the claim exceeds what small claims court can accommodate. If you have identified a template-level deficiency affecting thousands of accounts, if you are pursuing a demand in the tens or hundreds of thousands of dollars, if you are coordinating with regulatory filings across multiple agencies, then an attorney adds value through litigation infrastructure: drafting federal complaints, managing discovery in a way that maximizes disclosure, calculating damages across a complex statutory matrix, and negotiating settlements against corporate legal departments with deep resources.

In that scenario, the fee-shifting provisions flip from a tax into a weapon. The FDCPA’s mandatory attorney fee provision means the defendant pays your lawyer’s fees if you prevail. The Wisconsin Consumer Act provides the same. California’s § 1021.5 provides the same for any action vindicating public rights. The defendant is not just paying your damages—it is paying for the attorney who proved the damages. The fee-shifting provision was designed by Congress to make it economically rational for attorneys to take consumer protection cases they would otherwise decline, by guaranteeing that the violator—not the victim—bears the cost of enforcement.

But the critical point is this: the lawyer is the upgrade, not the prerequisite. You can enforce a binary compliance claim yourself, in your own name, in your own courthouse, with no legal training and no legal budget. The courthouse is open. The statute is public. The evidence is in your filing cabinet. You can win this case. Not “you might be able to win this case.” Not “with enough luck and preparation you could potentially prevail.” You can win. Because the claim is binary, the evidence is undisputed, and the law is on your side. The only question is whether you show up.

The Nightmare Compounds

Let us walk through what is actually happening inside the institutions that are the subject of a distributed compliance audit. Not the hypothetical version. The structural reality, right now.

Each discoverer is not just a plaintiff

A corporate general counsel receiving a single sophisticated demand letter from a single consumer has a playbook for that. Evaluate the claims. Assess the exposure. Decide whether to settle, litigate, or ignore. The decision tree is well-established and the variables are known. One adversary, one claim, one jurisdiction, one set of potential damages.

The distributed audit does not produce one adversary with a large claim. It produces an uncontrollable proliferation of independent evidentiary events.

Each independent discoverer is a potential plaintiff. They are also a potential regulatory complainant—to the CFPB, to their state attorney general, to the FTC, to the OCC, to a state department of financial institutions. They are a potential source for an investigative journalist. They are a potential witness in someone else’s litigation. They are a data point that a state attorney general can cite when justifying an investigation. They are a data point that a plaintiffs’ attorney can cite in a class certification motion. They are a social media post that goes viral and generates fifty more audits.

The general counsel cannot map this threat surface because the threat surface is definitionally unmappable. You cannot enumerate nodes in a distributed system when you do not know how many nodes exist, where they are, or when they will activate.

The notice problem multiplies exponentially

Every independent discovery is a separate knowledge event. This matters enormously in any legal framework where the institution’s knowledge of the violation affects liability, damages, or the availability of defenses.

When one person sends a demand letter identifying a template-level deficiency, the institution is on notice as of the date of receipt. When forty people in different states independently confirm the same deficiency and each files a complaint or sends a demand letter, the institution has forty independent records of notice, created by forty independent sources, documented in forty independent paper trails. The institution cannot settle quietly with one person to make the knowledge event disappear, because thirty-nine other knowledge events exist independently.

This is particularly devastating in the context of Sarbanes-Oxley certifications for publicly traded companies. When the CEO and CFO sign SOX §§ 302 and 906 certifications attesting to the completeness and accuracy of financial statements, their personal liability turns on what they knew and when they knew it. Each independent notice from each independent auditor creates an independent record that they were informed of a material contingent liability. The distributed audit does not just discover the noncompliance—it generates an expanding constellation of knowledge events that makes “we didn’t know” progressively less defensible with each passing week.

The settlement calculus collapses

Corporate legal departments settle claims based on a calculus: the cost of settlement versus the cost of litigation, adjusted for the probability of an adverse outcome and the reputational exposure of a public proceeding. This calculus assumes that settling a claim resolves it—that paying one plaintiff reduces the total threat surface by one.

In a distributed compliance audit, settling with one plaintiff does not reduce the threat surface at all. The next plaintiff does not know about the settlement. The next plaintiff does not care about the settlement. The next plaintiff has their own billing statement, their own AI analysis, and their own independent confirmation of the same template-level deficiency. Every settlement is an independent event that resolves one claim without affecting any other claim, because the claims are not coordinated.

Worse: settlement terms that include confidentiality provisions—the standard corporate tool for preventing one settlement from encouraging others—are useless against a distributed audit. The methodology is already public. The findings are independently reproducible. Silencing one plaintiff does not prevent the next plaintiff from independently discovering the same violation. The confidentiality provision protects the amount of the settlement but does nothing to suppress the finding that generated it.

And worse still: the first settlement functions as a pricing event. Not because the amount leaks—it may or may not—but because the fact of settlement signals that the institution concluded it was cheaper to pay than to defend. That signal, in an environment where the underlying finding is independently confirmable, is an invitation. Not a coordinated invitation. A structural one. The existence of a settlement tells every subsequent discoverer that the institution knows its templates are noncompliant and has already paid someone to go away rather than fix them.

Every settlement is an independent event that resolves one claim without affecting any other claim. The confidentiality provision protects the amount but does nothing to suppress the finding that generated it.

The General Counsel’s Impossible Position

Now let us put ourselves inside the general counsel’s office at an institution that has just learned it is the subject of a distributed compliance audit. Not a hypothetical institution. A real one: a billing platform that generates templates for tens of thousands of healthcare providers, or a hospital system that processes hundreds of thousands of patient accounts annually, or a bank whose health services financing program uses a standardized statement template across its entire portfolio.

The general counsel has just received a demand letter that identifies a template-level deficiency—missing mandatory Regulation F validation disclosures that violate federal law in every state where the template generated a statement, plus state-specific violations under whatever consumer protection framework applies in the complainant’s jurisdiction. In California, the missing disclosures render the contracts void under SB 1061—the debt becomes phantom debt, legally nonexistent. In Wisconsin, they generate independent violations under the Consumer Act and the DTPA with doubled damages. In Texas, knowing violations trigger treble damages. The demand letter is accompanied by individualized regulatory complaints to every agency with jurisdiction, a parallel filing with the CFPB, and a copy of a published essay that explains the methodology and instructs readers to check their own billing statements using AI.

What does the general counsel do?

Option one: fix the templates

The obvious response is to update the templates immediately. Add the missing disclosures. Achieve compliance going forward. This is the correct response and the one the statute intended to produce.

But fixing the templates going forward does not fix the communications that were already sent. Under SB 1061, the California contracts are void at formation with no cure provision—every dollar is phantom debt. But even outside California, every noncompliant statement already received by a patient in any state is an independent Regulation F violation with independent statutory damages. The FDCPA violations cannot be cured retroactively either—a noncompliant communication that was sent remains a noncompliant communication that was sent. Under Wisconsin’s Consumer Act, the false representation of the legal status of the debt has already occurred. Under Wisconsin’s DTPA, the deceptive representation was made at the moment the statement was received. Updating the template today does not un-send the statements from yesterday.

This means the general counsel must simultaneously fix the templates going forward and confront the question of what to do about the portfolio of noncompliant communications already sent. In California, the question is how many phantom debt contracts exist and what payments were collected on phantom obligations. In every other state, the question is how many noncompliant statements were sent, how many statutory violations they represent, and what the aggregate exposure looks like when you multiply per-violation damages across every affected account in every affected jurisdiction under every applicable statute. These questions require an internal audit of precisely the kind the institution has been avoiding, and conducting that audit generates discoverable documentation of the scope of noncompliance.

Option two: defend the templates

The general counsel could argue that the templates are compliant—that the mandatory disclosures are present, or that the statute does not apply, or that the AI’s analysis is wrong. This defense requires engaging with the substance of the finding.

The problem is that the finding is binary. The disclosure language is either on the document or it is not. This is not a legal interpretation that reasonable minds can dispute. It is a textual observation. The AI is not construing the statute. It is reading the document. Anyone—any attorney, any regulator, any judge—can read the same document and confirm whether the disclosure language is present. The defense collapses on first contact with the evidence because the evidence is the defendant’s own document.

Option three: ignore the demand

The general counsel could ignore the demand letter and hope the problem goes away. This strategy has historically been effective against individual consumer complaints, which tend to dissipate when the consumer runs out of energy or resources.

A distributed compliance audit does not dissipate. The methodology is public. The tool is available. New auditors are activating at a rate the institution cannot predict. Ignoring one demand letter does not prevent the next demand letter, the next regulatory complaint, the next social media post explaining what the auditor found. And every day of inaction after notice compounds the liability—because the institution is now knowingly collecting on phantom debt using communications it has been informed are noncompliant with federal law and with the state consumer protection statutes of every jurisdiction where those communications were received.

Option four: settle quietly and move on

As discussed above, settlement resolves one claim without affecting any other claim. Settling quietly with the first discoverer does not prevent the second, third, or fortieth discoverer from making the same finding independently. The distributed audit is immune to the standard corporate strategy of buying silence, because the silence of one auditor does not affect the existence or findings of any other auditor.

Option five: challenge the statute

The institution could mount constitutional or preemption challenges to state statutes—SB 1061 in California, the Consumer Act in Wisconsin, the DTPA in Texas—but this requires fighting separately in every state, with no guarantee of success in any of them, and with no effect whatsoever on the federal Regulation F and FDCPA violations that exist independently of every state statute. Defeating SB 1061 does not cure the Regulation F violations in Wisconsin. Defeating the Wisconsin Consumer Act claim does not cure the FDCPA claim. The federal violations are structural and universal. They cannot be challenged without arguing that the CFPB’s own disclosure requirements are unconstitutional—a position that is not only legally dubious but reputationally suicidal.

The Regulatory Prisoner’s Dilemma

The distributed compliance audit does not only create an ungovernable problem for the private institutions whose templates are noncompliant. It creates a parallel and equally ungovernable problem for every regulatory agency with jurisdiction over those institutions. This is by design.

The standard regulatory complaint goes to one agency. Maybe two. The complainant identifies the most relevant authority, files a complaint, and hopes for the best. The agency receives it, triages it, assigns it a case number, and either acts or does not. If it does not act, nobody else knows. The complaint disappears into a queue. The agency faces no consequences for inaction because no one is watching. This is how the system has always worked. It is also why the system has always failed.

What happens when the complainant does not file a single complaint but files individualized complaints to every agency with jurisdiction—simultaneously—and CCs every agency on every other agency’s complaint?

Not form letters. Not copies of the same document sent to different addresses. Individually tailored complaints, drafted to each agency’s specific statutory mandate, in each agency’s own regulatory language, addressing each agency’s own jurisdictional authority. The CFPB complaint speaks Regulation F. The state attorney general complaint speaks UDAP and the Rosenthal Act. The SEC filing speaks SOX certifications and material nonpublic information. The OCC complaint speaks bank supervision. The HHS Office for Civil Rights complaint speaks HIPAA. The FTC complaint speaks unfair and deceptive practices. The state department of financial institutions complaint speaks lending compliance. Each agency receives a document that was written specifically for it, about the specific statutes it is charged with enforcing, using the specific terminology its enforcement staff uses every day.

And then every single complaint includes a CC block listing every other agency that received a complaint about the same underlying conduct.

The structure of the dilemma

Game theory defines a prisoner’s dilemma as a situation in which each participant’s individually rational choice leads to a collectively suboptimal outcome. The classic formulation involves two prisoners who must independently decide whether to cooperate or defect, without knowing the other’s choice. The individually rational move for each prisoner is to defect. The collectively optimal outcome requires both to cooperate. The structure of the incentives guarantees the suboptimal result.

The regulatory prisoner’s dilemma inverts this structure. Here, the individually rational move for each agency is to act—and the collectively suboptimal outcome for the agencies is the one in which they all choose inaction. The structure of the incentives guarantees that at least one agency will move.

Here is why. Each agency that receives an individualized complaint knows the following things simultaneously:

First, that the complaint is substantive. It is not a form letter or a generic grievance. It is a detailed, statute-specific filing that identifies violations within that agency’s particular enforcement mandate. The agency’s staff can evaluate its merit on first reading because it is written in their language about their statutes.

Second, that every other agency with jurisdiction has received an equally substantive complaint about the same underlying conduct. The CC block is not decorative. It is an information forcing function. Every agency knows that every other agency knows.

Third, that the complainant has created a documentary record of notice to every agency simultaneously. This record exists regardless of what any agency does with it. It is available to Congress, to inspectors general, to journalists, to courts, and to the complainant in any subsequent litigation.

Fourth, that the agency’s response—or failure to respond—is visible to every other agency on the CC list and to the complainant who will document it.

Every agency knows that every other agency with jurisdiction has received a substantive complaint about the same conduct. Every agency knows that its response—or its silence—is visible to everyone else on the list.

Why inaction is the highest-risk strategy

In a standard regulatory complaint, the agency that does nothing faces no consequences. The complaint sits in a queue. Nobody monitors the queue from outside. If the agency never acts, the complainant may never know why and has no mechanism to impose accountability for the non-response.

The CC structure eliminates this asymmetry. When every agency with jurisdiction is simultaneously on notice, the agency that acts first captures the enforcement action—the investigation, the settlement, the press release, the demonstration of institutional competence. The agency that acts second can still participate, but it is following rather than leading. The agency that acts last faces questions about why it waited. And the agency that never acts faces a documentary record showing it received an individualized, substantive complaint within its statutory mandate and chose to do nothing while peer agencies moved.

That documentary record is not hypothetical. It is a certified mail receipt. It is a USPS tracking number. It is a portal submission confirmation with a timestamp. It exists in the physical world, in government archives, and in the complainant’s files. It cannot be deleted, altered, or made to disappear. It is the kind of record that inspectors general review, that congressional oversight committees subpoena, and that investigative journalists cite.

The agency that chooses inaction is betting that nobody will ever ask why it sat on a documented complaint while other agencies acted. In the current political environment—where federal consumer protection agencies are being scrutinized from both sides of the aisle for either overreach or dereliction—that is not a safe bet. An agency that received an individualized complaint, in its own statutory language, with evidence attached, and did nothing, is an agency that has handed its critics a ready-made example of institutional failure.

The dilemma compounds across jurisdictions

The structure becomes even more powerful when the complaints span both federal and state agencies across multiple states. A complaint to the Wisconsin Attorney General and a simultaneous complaint to the California Attorney General about the same entity creates a multi-state coordination problem for the respondent—but it also creates a multi-state visibility problem for the attorneys general themselves. If Wisconsin acts and California does not, California’s inaction is visible. If California acts and Wisconsin does not, Wisconsin’s inaction is visible. Neither AG can assume the other is handling it, because the complaints are tailored to each state’s specific statutes and each state’s specific consumer protection mandate.

The federal agencies face the same dynamic relative to each other and relative to the state agencies. The SEC cannot assume the CFPB will handle the Regulation F violations, because the CFPB complaint is about Regulation F and the SEC complaint is about SOX certifications and undisclosed contingent liabilities. The OCC cannot assume the state AG will handle the bank supervision issues, because the state AG complaint is about state consumer protection law and the OCC complaint is about federal banking regulation. Each complaint is scoped to the receiving agency’s specific authority, which means each agency knows that no other agency can fully address the complaint on its behalf.

The respondent’s general counsel, already facing the distributed compliance audit from the private sector, now faces a second distributed system: a regulatory apparatus in which multiple independent agencies with overlapping but distinct jurisdictions are simultaneously aware of the same underlying conduct, simultaneously in possession of substantive complaints tailored to their specific mandates, and simultaneously visible to each other. The general counsel cannot manage this through a single point of contact. There is no single point of contact. There are twelve.

The compounding effect on institutional defense

From the respondent’s perspective, the regulatory prisoner’s dilemma interacts catastrophically with the distributed compliance audit. The private-sector audit generates an expanding universe of independent complaints, demand letters, and evidentiary records from individual patients and auditors. The regulatory dilemma ensures that those complaints—plus the complainant’s own detailed, individualized agency filings—land simultaneously on the desks of agencies that are structurally incentivized to act rather than wait.

The general counsel now faces a multi-front engagement that no settlement strategy can consolidate. Settling with the original complainant does not resolve the regulatory complaints, which are now in the hands of independent agencies with independent enforcement authority. Settling the regulatory complaints—if that is even possible—does not resolve the distributed audit findings from other patients who are independently discovering the same template-level noncompliance. Settling everything simultaneously is impossible because the parties are not coordinated, do not know about each other, and have different interests.

The institution is not facing a lawsuit. It is facing a system—a distributed, multi-jurisdictional, self-reinforcing system in which private enforcement and regulatory enforcement amplify each other. The private auditors generate findings that strengthen the regulatory complaints. The regulatory activity generates public records that inform and encourage additional private auditors. The cycle has no natural stopping point except the one the statute always intended: compliance.

The Sorting Problem

There is a deeper structural crisis that the distributed audit exposes but does not create. The crisis was always there. The audit simply makes it visible.

Medical debt in the United States is not held exclusively by the entities that originated it. Medical receivables are packaged, sold, securitized, and distributed across the financial system through asset-backed securities, collateralized loan obligations, and institutional balance sheets. The face value of American medical debt is approximately $220 billion. A significant but unknowable percentage of that debt was originated through billing templates that are noncompliant with Regulation F and state consumer protection statutes. In California, noncompliance with SB 1061 renders the underlying contracts void—the debt is phantom debt, and phantom debt has exactly one value: zero. In every other state, the receivables carry undisclosed statutory damages exposure under the FDCPA and the applicable state consumer protection framework that no securitization model has priced. Whether the receivable is phantom debt with a literal value of zero or a receivable encumbered by unquantified contingent liabilities that dwarf its face value, the result for the capital structure is the same: the face value is not the real value. The $220 billion medical debt asset class is contaminated with phantom receivables, and no one in the securitization chain knows which ones they are.

The financial system does not know which receivables are phantom and which are real. There is no field in any securitization database that records whether the originating billing template contained the mandatory Regulation F disclosures, the SB 1061 language, or the state-specific disclosures required by the consumer protection statute of the patient’s home jurisdiction. There is no sorting mechanism that can separate phantom receivables from performing receivables within a securitized pool. The phantom receivables are commingled with the real receivables at every level of the capital structure—in the original portfolios, in the ABS tranches, in the CLO structures, on the institutional balance sheets that hold them.

This is the sorting problem. It is unsolvable from outside the originating institution, because only the originating institution knows what template was used to generate each contract. And it may be unsolvable from inside the originating institution, if the templates were not versioned or if compliance status was not tracked at the contract level—which, given that the templates predate many of the statutes, it almost certainly was not. The sorting problem is compounded by the all-fifty-states dimension: a single receivable originated in California has a different contamination profile than one originated in Wisconsin or Texas, because the applicable statutes and damages calculations differ by jurisdiction. A securitization pool containing receivables from multiple states contains multiple categories of contamination, each requiring separate analysis under separate legal frameworks.

The phantom receivables are commingled with the real receivables at every level of the capital structure. There is no sorting mechanism. The contamination spans all fifty states and the sorting problem is unsolvable.

Now consider what the distributed compliance audit does to this problem. Each independent auditor who confirms that a billing template is noncompliant is not just identifying a violation affecting their own account. They are producing evidence that the template generates noncompliant communications across every account that uses it. One noncompliant template means every statement generated by that template carries the same violations—whether those violations create phantom debt in California, generate triple-statute exposure in Wisconsin, or trigger treble damages in Texas. The auditor has identified not just their own noncompliant statement but the contamination of the entire portfolio originated through that template, in every state where it was used. Every receivable generated by that template is a phantom receivable.

When the portfolio is securitized, the phantom receivables propagate upward through the capital structure. The ABS tranche that contains phantom receivables is worth less than its face value—how much less depends on the jurisdictional mix and the applicable damages calculations, which no one in the securitization chain has modeled because no one knew the phantom receivables existed. The CLO structure that holds that tranche is worth less than its rated value. The institutional balance sheet that carries the CLO position is overstated. The lending capacity that depends on that balance sheet is built on phantom assets. Phantom debt at the origination level becomes phantom capital at the institutional level.

I coined the term phantom debt to describe this phenomenon, and phantom receivables to describe what these debts become when they enter the financial system. Phantom debt is debt that appears real on balance sheets, generates revenue, gets collected and securitized, but is legally nonexistent or legally encumbered because the originating contracts are void or the originating communications violate federal and state law. Phantom receivables are what phantom debt becomes when it is packaged into asset-backed securities and distributed across the capital structure. In California, where SB 1061 applies, phantom receivables have exactly one value: zero. No floor, no cure, no recovery. In every other state, the phantom receivable’s face value is offset by undisclosed contingent liabilities—statutory damages, doubled or trebled damages, mandatory attorney fees—that no balance sheet reflects. The phantom receivable looks real. It performs like a real asset. It generates cash flows. But the cash flows are extracted from people who do not owe the money, through communications that violate the law, on debts that are either legally nonexistent or legally indefensible. Phantom debt is the largest undisclosed systemic risk in American consumer finance, and the distributed compliance audit is making it visible for the first time.

The distributed compliance audit does not create the phantom receivables. They were created the moment the noncompliant template generated the first statement. What the distributed audit does is make their existence discoverable, provable, and impossible to ignore—in California, in Wisconsin, in Texas, in every state where the template was used—because the discovery methodology is public, the findings are independently reproducible, and the evidence is the institutions’ own documents.

The Expected Cost Calculation Has Permanently Changed

Every corporate risk model for billing compliance rests on an expected cost calculation: the probability of detection multiplied by the cost if detected. This calculation has historically favored noncompliance because the probability of detection was near zero.

The distributed compliance audit changes this calculation permanently. Not by changing the cost of detection—though it does that too, reducing it from thousands of dollars to zero—but by changing the probability of detection. When the methodology for identifying noncompliance is published, freely available, and executable by any person with a phone and a twenty-dollar subscription, the probability that a given noncompliant template will be detected is no longer near zero. It is a function of time, the number of patients who received statements generated by that template, and the rate at which those patients encounter the methodology.

The rate of encounter is the variable that institutions cannot control and cannot model. It depends on how many people read the methodology. How many of them act on it. How many of them share their findings. How many of those findings generate media coverage, social media engagement, regulatory action, or additional demand letters. The rate is not constant. It is self-amplifying: each discovery that generates a visible outcome increases the probability that other people will try the methodology on their own documents.

Over time—and not a long time—the probability of detection for any given noncompliant template approaches certainty. This is not because the distributed audit is omniscient. It is because the population of potential auditors is the entire population of patients who received statements generated by the template, and the cost of auditing is zero. In a population of thousands or tens of thousands of affected patients, the probability that none of them will ever upload their statement to an AI and ask the question is vanishingly small. It requires sustained universal ignorance of a methodology that has been published and is actively spreading.

The probability of detection for any given noncompliant template approaches certainty over time. It requires sustained universal ignorance of a methodology that has been published and is actively spreading. That is not a defensible assumption.

This means the expected cost of noncompliance is converging toward the full cost of the violation—statutory damages, refund exposure, regulatory penalties, litigation costs, and reputational harm—multiplied by the number of affected accounts. The expected cost calculation that made noncompliance economically rational has inverted. From this point forward, noncompliance is not cheaper than compliance. It is catastrophically more expensive.

What Cannot Be Stopped

The corporate defense toolkit for managing regulatory and litigation risk includes a well-established set of instruments: settlement with confidentiality, lobbying for legislative reform, challenging the underlying statute, preemption arguments, arbitration clauses, standing challenges, motion practice to increase the cost of litigation, public relations campaigns, and institutional capture of enforcement agencies. Every one of these instruments assumes a finite number of adversaries operating through identifiable channels with negotiable interests.

The distributed compliance audit has none of these properties. It is not a party that can be settled with. It is not an organization that can be lobbied against. It is not a legal theory that can be challenged. It is not a plaintiff whose standing can be questioned. It is a factual observation—that the disclosure is missing from the document—being independently confirmed by an unknowable number of independent observers using independently available tools.

You cannot enjoin someone from reading their own billing statement. You cannot obtain a protective order against a person asking an AI whether a document is compliant with public law. You cannot file a motion to dismiss a regulatory complaint before it generates an investigation. You cannot settle with a person who has not yet discovered the violation but will discover it next week, or next month, or next year.

The distributed compliance audit is immune to the standard corporate defense toolkit because the toolkit was designed to manage disputes between identifiable parties, and the distributed audit is not a dispute. It is a condition. The noncompliance exists in the documents. The methodology for detecting it exists in the public domain. The tools for executing the methodology exist on every smartphone. The condition persists until the documents are changed.

This is the part that should be mortifying for every corporate officer who has relied on the traditional risk calculus. Not that they are being audited. Not that the violations are being discovered. Not even that the liability is enormous. What should be mortifying is the realization that the only way to make the distributed compliance audit stop generating adverse findings is to actually become compliant.

That has always been the answer. The statutes have always required these disclosures. The compliance obligation has always existed. The only thing that changed is that the probability of being caught for ignoring it went from near zero to near one.

The law did not change. The math changed.

Industrial-Scale Enforcement From a Kitchen Table

Everything described in this essay so far has assumed the distributed audit operates at the pace of individual discovery—one person, one billing statement, one finding at a time. That assumption is already obsolete. The same AI tools that enable detection also enable prosecution at a scale that no individual litigant has ever achieved and that no corporate legal department is prepared to confront.

Consider what the production of a demand letter actually requires. You must identify the entity. You must identify the specific statutes it has violated—federal and state, with the correct section numbers and the correct mandatory disclosure requirements for each. You must identify the entity’s officers by name. You must locate their addresses—registered agent, headquarters, legal department, and in some cases personal residences for service of notice. You must calculate damages: per-violation statutory damages under each applicable statute, actual damages, and where applicable, multipliers—doubled damages under Wisconsin’s DTPA, treble damages under RICO, treble damages under the Texas DTPA. You must draft the letter in a register that demonstrates command of the legal framework—not a form letter, but a bespoke document tailored to that specific entity, that specific set of violations, and that specific damages calculation. You must prepare the CC block listing every regulatory agency that will receive a parallel complaint. You must format the document for certified mail. And you must do this for every entity in your portfolio.

Five years ago, this process would have consumed weeks of attorney time per entity. The research alone—identifying the applicable statutes, locating the officers, calculating the damages—would require a paralegal and a Westlaw subscription. The drafting would require an attorney who could write in the register that corporate legal departments take seriously. The production of a single demand letter was a professional service that cost hundreds or thousands of dollars. Building a portfolio of cases against a dozen entities simultaneously was the work of a litigation team at a mid-size law firm, billed at rates that no individual plaintiff could afford.

I built a portfolio of cases against over a dozen entities in a single week. Every demand letter individually tailored. Every damages calculation independently derived. Every CC block listing every relevant regulatory agency. Every document formatted for certified mail, ready to print and sign. Not because I have a litigation team. Not because I have a law degree. Because I have an AI that functions as a bespoke legal document production engine and a parallel processing pipeline that generates finished, formatted documents simultaneously across every entity in the portfolio.

The AI does not produce form letters. It produces individually crafted documents that address each entity’s specific violations under each entity’s specific statutory exposure. The demand letter to a California hospital speaks SB 1061, the Rosenthal Act, UCL § 17200, and HIPAA. The demand letter to a Wisconsin health system speaks the Wisconsin Consumer Act, § 100.18, and Regulation F. The demand letter to a Texas debt collector speaks the Texas DTPA with treble damages. The SEC whistleblower filing speaks SOX certifications, MNPI, and cross-portfolio co-conspiracy. Each document is built from the same underlying factual findings—the same template-level noncompliance—but translated into the specific legal language that applies to that specific recipient in that specific jurisdiction. This is not mail merge. This is industrial-scale bespoke legal production.

I built a portfolio of cases against over a dozen entities in a single week. Every demand letter individually tailored. Every damages calculation independently derived. This is not mail merge. This is industrial-scale bespoke legal production.

The parallel production capability is the part that should terrify corporate legal departments. I am not sitting at a desk looking up addresses one at a time. I am not manually researching which statutes apply in which jurisdiction. I am not spending days formatting documents. The AI knows the statutes. The AI calculates the damages. The AI generates the documents. The parallel pipeline produces them simultaneously—not sequentially, not one at a time over weeks, but all at once, formatted and ready for wet-ink signature. The bottleneck is not research, not drafting, not formatting. The bottleneck is how fast the post office can process certified mail.

This means a single person can blitzkrieg an entire corporate legal department with simultaneous individually tailored certified mail. Not one letter that the general counsel can triage and assign to an associate. A dozen letters arriving simultaneously, each one addressing a different entity, each one demonstrating command of the specific statutory framework applicable to that entity, each one identifying the phantom debt in that entity’s portfolio and quantifying the phantom receivables on its balance sheet, each one accompanied by regulatory complaints filed simultaneously with every relevant agency. The general counsel does not get to address these sequentially. They arrive at once. The knowledge events they create are simultaneous. The response deadlines they generate are concurrent. The institutional bandwidth required to address them exceeds the capacity of any legal department that did not know they were coming.

The economics of enforcement have inverted

Think about what this means for the cost structure of private enforcement. A plaintiff-side consumer protection law firm that takes on a case against a single entity commits an attorney, a paralegal, and months of calendar time. The firm’s capacity is constrained by headcount. It can handle a finite number of matters simultaneously, and each new matter requires incremental human labor.

An individual with an AI production pipeline operates on entirely different economics. The marginal cost of adding a new entity to the portfolio is near zero. The research is instant. The drafting is instant. The damages calculation is instant. The document production is instant. The only costs that scale linearly are postage and certified mail fees. A person who can prosecute one entity can prosecute twelve entities with approximately the same effort—because the effort is not in the production. The effort is in the original analytical insight: identifying the template-level noncompliance and understanding the statutory framework that makes it actionable. Once that insight exists, the AI replicates it across every entity it touches, in every jurisdiction those entities operate, under every statute that applies.

The law firm bills $500 an hour and produces one demand letter per week. The citizen enforcer pays $20 a month and produces a dozen in the same week, each one tailored to a different entity under a different combination of federal and state statutes. The law firm’s output is constrained by the number of attorneys on staff. The citizen enforcer’s output is constrained by nothing except the scope of the noncompliance and the number of entities involved.

This is not a hypothetical future capability. This is what I did. From Waukesha, Wisconsin. With a Finance MBA and a $20 AI subscription. While the CFPB was being dismantled and the FTC was being gutted and state attorneys general were being overwhelmed. One person, operating at industrial scale, generating individually tailored legal documents across more than a dozen entities simultaneously, each one creating independent knowledge events, independent regulatory complaints, and independent settlement demands—each one identifying phantom debt, naming phantom receivables, and quantifying the phantom capital that the recipients have been booking as real assets. Binary factual findings that the recipients cannot dispute, arriving simultaneously, from a single person they have never met, in a jurisdiction a thousand miles from their headquarters.

The Obligation of the Auditor

I want to close by speaking directly to the people who are performing these audits—the readers who took the call to action, uploaded their billing statements, and found exactly what I said they would find.

You are not doing something trivial. You are performing a function that the government was supposed to perform and is no longer performing. The CFPB has been gutted. The FTC is being dismantled. State attorneys general are overwhelmed. The enforcement infrastructure that was supposed to protect you from exactly this kind of noncompliance has been systematically destroyed. You are filling the gap, one billing statement at a time, using tools that did not exist two years ago.

The private attorney general doctrine—the legal framework that makes all of this possible—carries obligations. You are not just finding violations. You are generating evidence. That evidence has value to regulators, to other patients, to the legal system. Treat it accordingly.

Document what you find. Keep the original billing statements. Screenshot the AI analysis. Note the date you performed the audit. If you file a regulatory complaint, retain the confirmation. If you send a demand letter, send it by certified mail. Create a paper trail that can be independently verified, because the strength of the distributed audit depends not on any individual finding but on the aggregate weight of independently documented findings pointing to the same conclusion.

You do not need a law degree to do this. You do not need a litigation budget. You do not need anyone’s permission. You do not need to live in California. The federal statutes that protect you apply in every state. Your state’s consumer protection laws add independent remedies on top of the federal floor. The tools that enable the audit cost twenty dollars a month. The evidence is in your filing cabinet. The rights are yours. The methodology works. I know it works because I used it from Waukesha, Wisconsin—not San Francisco, not New York, not any jurisdiction that anyone would have predicted—and over a dozen entities that thought they could collect on phantom debt are currently confronting the consequences of individually tailored demand letters, regulatory complaints, and SEC whistleblower filings that arrived simultaneously and that they cannot make go away.

You are the distributed compliance audit. You are the mechanism the system did not account for. You are the reason the expected cost calculation is changing, institution by institution, template by template, billing statement by billing statement.

Keep going.

Andrex Ibiza

Waukesha, Wisconsin

March 2026

Private Enforcement, the Rise of the Judge, and the Future of American Consumer Protection

In 2024, I had surgery. Afterward, I received billing statements from nine separate medical entities. The statements arrived on schedule, formatted professionally, bearing account numbers and balances due and minimum payment amounts. They looked the way medical bills have always looked in America—authoritative, final, non-negotiable. I was expected to pay them. Over one hundred million Americans receive statements like these and do exactly that.

I asked a different question. I uploaded one of the statements to an artificial intelligence tool—Anthropic’s Claude—and asked whether it contained all the disclosures required by law. The answer came back in seconds: no.

I checked the next statement. Also no. I checked all nine. Nine entities. Nine sets of billing communications. Zero compliant with the federal and state statutes governing medical debt collection in the United States. Mandatory validation disclosures required by the Consumer Financial Protection Bureau’s Regulation F were absent. Original creditor identification was missing. Dispute notices were nowhere to be found. In California, where my surgery took place, a new statute—Civil Code § 1785.27, enacted as SB 1061—had rendered any medical debt contract missing specific credit reporting prohibition disclosures void and unenforceable by operation of law, effective July 1, 2025. No cure provision. No grace period. No substantial compliance defense.

I entered that process believing I had approximately ten thousand dollars in medical debt. I exited it understanding that my medical debt did not legally exist—not reduced, not disputed, not negotiated, but nonexistent by operation of law.

What followed was a campaign of demand letters, regulatory complaints, and federal filings that is still ongoing. But the implications of what I discovered extend far beyond my own billing dispute. They reach into the foundations of American consumer protection law, the infrastructure of medical debt collection, and the question of who enforces the law when the government will not. This paper is an attempt to synthesize those implications into a single framework.

Part I: The Private Attorney General

The invention of the citizen enforcer

The United States does not enforce its public laws the way most democracies do. Other nations maintain large regulatory bureaucracies staffed with inspectors, investigators, and prosecutors who detect violations and bring enforcement actions on behalf of the state. The American system relies on something fundamentally different: it deputizes private citizens to enforce public law through private litigation, then compensates them with attorney fee awards funded by the violators they catch. The legal name for this mechanism is the private attorney general doctrine. It is the single most important structural feature of the American legal system that most Americans have never heard of.

The doctrine’s intellectual origin traces to Judge Jerome Frank of the Second Circuit, who in Associated Industries of New York State v. Ickes in 1943 coined the phrase “private Attorney Generals” to describe citizens Congress authorizes to sue in the public interest. For three decades, the concept remained academic. Then a barbecue chain in South Carolina refused to serve Black customers, and the resulting Supreme Court case—Newman v. Piggie Park Enterprises in 1968—established the rule that still governs: a plaintiff who successfully enforces civil rights law acts not for himself alone but as a private attorney general vindicating a policy Congress considered of the highest priority. Successful plaintiffs should ordinarily recover attorney fees unless special circumstances render the award unjust.

This was revolutionary because it broke from the American Rule—the longstanding principle that each party pays its own attorney fees regardless of outcome. Before Piggie Park, fee-shifting existed only through narrow common-law exceptions. The decision grafted a new, purpose-driven exception onto the framework and opened the door for Congress to create statutory fee-shifting provisions across virtually every area of regulatory law.

The crisis and resurrection of 1975–1976

The doctrine’s most dramatic chapter unfolded in eighteen months. In Alyeska Pipeline Service Co. v. Wilderness Society in 1975, the Supreme Court held that federal courts lack inherent equitable power to create fee-shifting exceptions to the American Rule. Only Congress can do that. The backlash was bipartisan and immediate. Within eighteen months, Congress passed the Civil Rights Attorney’s Fees Awards Act of 1976, explicitly invoking Piggie Park and declaring that plaintiffs in civil rights cases act as private attorneys general whose work benefits all citizens.

Congress has since enacted fee-shifting provisions in over two hundred federal statutes and more than two thousand exist at the state level. The Fair Debt Collection Practices Act. The Clean Air Act. The Americans with Disabilities Act. The Endangered Species Act. The Truth in Lending Act. Each one converts private litigants into enforcement agents, making it economically viable to bring suits that vindicate public rights. Without these provisions, most civil rights, environmental, and consumer protection litigation in America would be financially impossible.

The asymmetric design

A defining feature of the doctrine is its deliberate one-way ratchet. In Christiansburg Garment Co. v. EEOC in 1978, the Supreme Court established that prevailing civil rights plaintiffs should ordinarily recover fees, while prevailing defendants may recover only when the plaintiff’s claims were frivolous, unreasonable, or without foundation. The rationale is explicit: the plaintiff is the chosen instrument of Congress to vindicate a policy of the highest priority, and symmetric fee-shifting would chill meritorious claims.

When fees are awarded, courts use the lodestar method from Hensley v. Eckerhart in 1983: hours reasonably expended multiplied by a reasonable hourly rate. The resulting figure is presumptively reasonable. Fee awards need not be proportionate to damages recovered—the Court in City of Riverside v. Rivera upheld a fee award of $245,456 when the underlying damages were far smaller. The message is structural: the value of enforcement is independent of the value of any individual plaintiff’s claim.

California’s broader vision

While the federal doctrine depends entirely on statute-specific Congressional authorization, California’s Code of Civil Procedure § 1021.5 operates as a general-purpose fee-shifting provision—arguably the most expansive in the nation. Enacted in 1978, it allows fee awards in any action that enforces an important right affecting the public interest, confers a significant benefit on the general public, and where the necessity and financial burden of private enforcement make the award appropriate. It does not require an underlying fee-shifting statute. Any claim vindicating public rights can qualify.

California also retains the catalyst theory—the principle that a plaintiff whose lawsuit causes voluntary reform can recover fees even without a judgment—which the U.S. Supreme Court rejected for federal claims in Buckhannon Board & Care Home v. West Virginia in 2001. The California Supreme Court found the federal reasoning unpersuasive. This divergence matters enormously: California’s framework incentivizes enforcement even when defendants capitulate, while the federal framework creates what empirical research has identified as strategic capitulation—defendants changing behavior without formal resolution specifically to avoid fee liability.

California’s SB 1061, effective in stages beginning January 1, 2025, represents the latest and most powerful expression of this philosophy. By rendering noncompliant medical debt contracts void and unenforceable by operation of law—with no cure provision—the statute creates a private enforcement mechanism that does not depend on any government agency to function. The law enforces itself at the moment of contract formation. Citizens who discover noncompliance need only point to the statute. The contract was never valid. There is nothing to litigate.

The statute executes automatically. The law enforces itself at the moment of contract formation. Citizens who discover noncompliance need only point to the statute.

Part II: The Collapse

The CFPB has been dismantled

Federal consumer protection enforcement in the United States has undergone the most dramatic retrenchment in modern history. The Consumer Financial Protection Bureau entered January 2025 with 1,755 employees and 34 active enforcement actions. By year’s end, 88% of the workforce had been targeted for elimination. Sixteen of 34 pending enforcement actions were dismissed with prejudice—meaning they can never be refiled. Only one new enforcement action was filed in all of 2025.

Russell Vought, a Project 2025 architect with no consumer finance background, was installed as Acting Director on January 31, 2025, after Director Rohit Chopra was fired. Within days, Vought issued stop-work orders, closed headquarters, and told employees not to perform any work tasks without prior approval. He stated publicly that his goal was to shut the agency down within months.

The regulatory rollbacks were equally sweeping. The credit card late fee cap—projected to save consumers ten billion dollars annually—was vacated. The overdraft fee rule, projected to save five billion per year, was overturned through the Congressional Review Act, permanently barring any substantially similar regulation. The medical debt credit reporting rule, which would have removed forty-nine billion dollars in medical bills from credit reports for fifteen million Americans, was vacated. Sixty-seven guidance documents spanning every major consumer financial law were rescinded in a single day. The budget was cut nearly in half. The acting head of enforcement resigned, writing that the bureau’s leadership had no intention to enforce the law in any meaningful way.

An estimated nineteen billion dollars in annual consumer protections has been stripped away.

The FTC faces its own existential crisis

The Federal Trade Commission’s trajectory points in the same direction. The new chair arrived with a leaked pitch document promising to be more receptive to mergers and to stop legally dubious consumer protection cases. The President fired both Democratic commissioners in violation of the FTC’s governing statute and ninety years of Supreme Court precedent. The resulting case reached the Supreme Court, where the Chief Justice signaled during oral arguments that the conservative majority may overturn the foundational principle of independent agency governance. The FTC currently operates with two commissioners, both Republican—the bare minimum for a quorum.

The agency is implementing a 15% workforce reduction to its lowest staffing level in a decade. The Bureau of Consumer Protection is losing 32 full-time positions and eighteen million dollars. The Click-to-Cancel rule was struck down. The commercial surveillance rulemaking was abandoned. Novel enforcement theories involving AI discrimination, data privacy, and individual executive accountability have been shelved.

State attorneys general are overwhelmed

Twenty-two state attorneys general filed suit to preserve the CFPB’s operations. A 42-state bipartisan coalition formed to address AI-related consumer risks. The Democratic Attorneys General Association launched a working group with former Director Chopra as senior advisor. Individual states have achieved notable results—Massachusetts secured the largest civil penalty order in state history; New York enacted its most significant consumer protection update in 45 years; Texas brought the first AI healthcare enforcement action.

But the structural limitations are severe. State attorneys general lack the CFPB’s funding mechanism, can only act within their jurisdictions, and depend on the CFPB’s consumer complaint database for investigations—a database whose complaint resolution rate has fallen from 50% to under 5%, with over 830,000 complaints currently unresolved. Enforcement has become starkly partisan. The federal enforcement vacuum is not being filled. It is being acknowledged, worked around, and endured.

Medical debt: a crisis without an enforcer

Over one hundred million Americans carry medical debt—41% of adults—totaling at least $220 billion. Approximately fifteen million Americans have medical debt on their credit reports. The burden falls disproportionately on Black Americans, residents of non-Medicaid-expansion states, and those with disabilities. Industry analyses estimate that 49% to 80% of medical bills contain at least one error, with hospital bills over $10,000 averaging $1,300 in errors. The FBI estimates fraudulent billing practices represent 3% to 10% of total health spending.

The CFPB’s medical debt credit reporting rule is dead. Its medical debt collection advisory opinion has been withdrawn. The No Surprises Act has produced only $4.18 million in monetary relief against a $220 billion problem. The independent dispute resolution process is dominated by private-equity-backed physician staffing companies. The agency that was supposed to protect these hundred million Americans has been functionally eliminated.

The National Consumer Law Center’s survey of 117 consumer advocates found that debt collectors generally do not comply with requests from consumers to cease contacting them and continue to sue on time-barred debts despite Regulation F’s prohibition. The CFPB has brought no standalone enforcement action for Regulation F violations. The entire enforcement burden has shifted to private litigants.

The entire enforcement burden has shifted to private litigants—and about 44% of FDCPA plaintiffs are repeat filers, indicating a small cadre of judges carrying the weight for everyone else.

Part III: The Convergence

The doctrine meets the vacuum

The private attorney general doctrine was designed for exactly this moment. Not as an emergency measure or a workaround, but as the foundational enforcement mechanism of the American legal system—the mechanism that two hundred federal statutes and two thousand state statutes were built to support. When Congress enacted fee-shifting provisions in the FDCPA, the FCRA, the TCPA, and dozens of other consumer protection statutes, it was making a structural bet: that private citizens, incentivized by fee awards and statutory damages, would detect and punish violations that government agencies could not catch on their own.

The bet was not supposed to be tested in isolation. The private attorney general was designed to operate alongside federal agencies with supervision authority, subpoena power, rulemaking capacity, and billion-dollar budgets. The doctrine was the complement, not the substitute. But the agencies have been destroyed, and the doctrine is what remains.

The data confirms the shift is already happening. According to WebRecon’s 2025 annual data, FCRA lawsuits reached 8,369 filings—up 37.4% from the prior year. FDCPA filings rose 7.8% to 4,534. TCPA class actions hit 1,052 in the first half of 2025 alone—a 95% increase year over year. CFPB complaints against debt collectors nearly doubled, from 159,732 in 2024 to 302,133 in 2025. As federal enforcement has collapsed, private enforcement has surged. The inverse relationship is not a coincidence. It is the system working as designed—under stress conditions it was never intended to bear alone.

The obstacles are structural

The Supreme Court has spent decades narrowing the private attorney general’s operating space. Buckhannon eliminated the catalyst theory for federal claims, encouraging strategic capitulation. City of Burlington v. Dague prohibited contingency risk multipliers, creating a structural impediment to securing representation. Lackey v. Stinnie held that even preliminary injunctions do not confer prevailing party status. TransUnion v. Ramirez narrowed standing for consumer protection claims. Evans v. Jeff D. permits defendants to condition settlement on waiver of statutory attorney fees, creating a conflict between the plaintiff’s interest in relief and their attorney’s interest in compensation.

Forced arbitration provisions cause an estimated 98% of cases that would otherwise be brought in some forum to be abandoned. Only 1 in 10,400 workers subject to arbitration has filed a claim. The access-to-justice crisis is severe: 92% of civil legal problems reported by low-income Americans receive no or inadequate legal help, and 120 million legal problems go unresolved annually. More than 75% of civil cases have at least one self-represented party. Pro se plaintiffs historically win only 4% of the time.

Legal scholar Myriam Gilles has documented that corporate lobbying campaigns have successfully narrowed consumer enforcement mechanisms in state legislatures. Professor Nancy Leong has shown that restrictive fee-shifting decisions have limited lawyers’ willingness to represent civil rights plaintiffs. The private attorney general is being asked to carry an unprecedented enforcement burden with one hand tied behind its back by the same judiciary that is supposed to facilitate its work.

The self-executing statute as a design breakthrough

Against this backdrop, California’s SB 1061 represents something new: a statute that does not depend on enforcement at all. It does not require an agency to investigate. It does not require a court to adjudicate. It does not require a plaintiff to prevail. The statute renders noncompliant contracts void at formation. The moment a medical debt contract is created without the mandatory disclosure language, it ceases to exist as a legal instrument. The patient does not need to know this. The provider does not need to know this. The contract is void whether or not anyone ever examines it.

This is a fundamentally different approach to consumer protection. Traditional enforcement requires detection, investigation, prosecution, and adjudication—a multi-year, multi-million-dollar process that depends on institutional capacity. Self-executing statutes require nothing. They operate automatically, continuously, and universally. The enforcement mechanism is the law itself.

The CFPB’s own January 2025 report—published in the final days of the Chopra directorship—explicitly recommended that states create representational or qui tam causes of action and authorize forms of private enforcement that can remain viable in the face of forced arbitration. SB 1061 goes further than even that recommendation. It makes enforcement unnecessary by making noncompliance self-punishing. You do not need to sue someone for breach of a contract that never existed.

Self-executing statutes require nothing. They operate automatically, continuously, and universally. The enforcement mechanism is the law itself.

Part IV: The Judge

Artificial intelligence changes the equation

The access-to-justice crisis in the United States is not a problem of law. The statutes exist. The rights exist. The fee-shifting provisions exist. The problem is that the people the statutes are supposed to protect cannot access them. They cannot afford attorneys. They cannot navigate complex legal procedures. They cannot identify when their rights have been violated because the violations are embedded in opaque billing systems, dense contracts, and institutional processes designed to discourage scrutiny.

Artificial intelligence eliminates the knowledge barrier. A consumer who uploads a billing statement to a frontier AI model and asks whether it contains all legally required disclosures receives an answer in seconds. That answer—which would have required a consumer protection attorney, a retainer fee, and weeks of analysis five years ago—is now available to anyone with a smartphone and a twenty-dollar monthly subscription. The cost of detecting a violation has collapsed from thousands of dollars to zero.

The evidence, though still early, is striking. A consumer in Long Beach, California, facing eviction after losing a jury trial with an attorney, used AI tools costing forty dollars per month combined to represent herself on appeal. She overturned the eviction and avoided approximately $73,000 in penalties. Another consumer reduced a $195,000 hospital bill by 83% using AI to identify duplicate charges, billing errors, and inflated costs. A May 2025 survey of 112 legal aid professionals found that 88% see AI as a tool to address the justice gap, and legal aid organizations are adopting AI at twice the rate of other lawyers.

The legal AI market is projected to grow from $3.11 billion in 2025 to $10.82 billion by 2030. States including Arizona, Utah, and Washington have reformed legal services regulations to allow new models of delivery. The ABA issued its first formal ethics opinion on generative AI in July 2024. The infrastructure is being built in real time.

The democratization of compliance analysis

What AI does to consumer protection enforcement is analogous to what the printing press did to religious authority. Before Gutenberg, the Bible existed in Latin, accessible only to clergy who could interpret it for the laity. The printing press put the text directly in people’s hands, in their own language, and the interpretive monopoly collapsed. Before frontier AI, consumer protection statutes existed in legal databases, accessible only to attorneys who could interpret them for clients. AI puts the analysis directly in people’s hands, in plain language, and the compliance-analysis monopoly is collapsing.

This matters because the medical billing system in the United States operates on an information asymmetry so extreme it resembles a feudal arrangement. The provider knows what was billed. The insurer knows what was approved. The billing platform knows what disclosures were included. The patient knows nothing except the amount due. Every entity in the chain has more information than the person being asked to pay, and the billing communications are designed—structurally, architecturally, at the template level—to maintain that asymmetry.

AI collapses the asymmetry. A patient who can ask an AI to compare their billing statement to the statutory requirements is no longer dependent on the provider, the insurer, or the billing platform to tell them whether their rights have been respected. They can check for themselves. And when they check, they may discover—as I did—that the documents in their hands are noncompliant with the statutes that govern them.

The replication problem for institutions

The most destabilizing implication of AI-powered compliance analysis is that the discovery mechanism is trivially replicable. A violation identified by one consumer using one AI tool can be independently verified by any other consumer using any other AI tool asking the same question about the same category of document. The discovery is not proprietary. It is not a trade secret. It is not protected by any barrier to entry. It is a question and an answer, and both are obvious to anyone who thinks to ask.

This creates an existential problem for any institution whose business model depends on consumer ignorance of their rights. When the cost of detecting a violation was thousands of dollars—an attorney’s retainer, a legal consultation, hours of research—the practical barriers to enforcement were high enough that widespread noncompliance was economically rational. The expected cost of being caught was low because so few consumers had the resources to check. The regulatory agencies were supposed to fill that gap, but even at full staffing they could only audit a fraction of the market.

AI changes the expected cost calculation permanently. When the cost of detecting a violation is zero and the tool is available to every person with a phone, the probability that noncompliance will be detected approaches certainty over time. Every billing statement that omits a required disclosure is a liability waiting to be identified by anyone who thinks to ask whether it’s compliant. The question is not whether it will be identified. The question is when.

When the cost of detecting a violation is zero and the tool is available to every person with a phone, the probability that noncompliance will be detected approaches certainty over time.

From individual remedy to systemic discovery

There is a difference between using AI to dispute your own bill and using AI to identify systemic noncompliance across an industry. The first is a consumer self-help tool. The second is an investigative methodology with implications that scale beyond any individual dispute.

When a consumer identifies that their billing statement is missing mandatory disclosures, the natural next question is whether the omission is specific to their account or embedded in the template that generates statements for all accounts. If it is a template-level deficiency—as it often is, because billing platforms generate standardized communications across thousands or millions of accounts—then the individual consumer’s discovery is evidence of portfolio-wide noncompliance. One person checking one statement can identify a violation affecting every patient who received a statement generated by the same template.

This is the scalability mechanism that institutions have not accounted for. Their compliance failures were tolerable when the probability of detection was low and the scope of any individual discovery was limited. AI changes both variables simultaneously: it makes detection trivially easy and it enables the discoverer to recognize the systemic nature of the violation immediately, because the AI can explain not just that the disclosure is missing but why it’s missing, what statute requires it, and what the consequences of the omission are across the entire population of affected accounts.

The consumer who discovers a template-level violation is not just a patient with a billing dispute. They are something the system was not prepared for: a judge. Not a judge appointed by the government or confirmed by the Senate or seated in a courtroom. A judge in the original and most fundamental sense—a person who examines the evidence, applies the law, and delivers the verdict. The private attorney general doctrine gave them the authority. The self-executing statute gave them the mechanism. The AI gave them the sight. They are acting not for themselves alone but for every person affected by the same noncompliant template. And the statutory framework—FDCPA damages, Regulation F violations, state UDAP claims, SB 1061 void-contract provisions—is already in place to enforce the judgment.

Part V: The New Paradigm

What the judge looks like in 2026

The traditional image of the private attorney general is an attorney—typically at a public interest law firm or a plaintiff-side consumer protection practice—who takes cases on a contingency or statutory-fee basis and brings enforcement actions that government agencies are unable or unwilling to pursue. That model still exists, but it is no longer the only model and it may no longer be the most important one.

The new figure is the judge: a person with no law degree, no institutional affiliation, and no litigation budget who uses commercially available AI tools to identify statutory violations in documents they encounter in the course of their daily lives, then leverages the existing statutory framework—fee-shifting provisions, statutory damages, void-contract remedies, regulatory complaint mechanisms—to deliver the verdict the system failed to deliver on its own. The tools cost twenty dollars a month. The statutes are public law. The evidence is the defendant’s own billing statements. The enforcement mechanism is already built. The judge does not ask permission. The judge reads the law, examines the evidence, and acts.

This is not hypothetical. It is happening now. FCRA filings are up 37%. FDCPA filings are up 8%. TCPA class actions are up 95%. The surge in private enforcement corresponds precisely with the collapse of federal enforcement and the proliferation of AI tools. The judge is not a future possibility. It is a present reality generating measurable results in federal courts.

The structural advantages of decentralized enforcement

Professor Myriam Gilles has argued that deploying citizens to enforce consumer protection laws generates a more stable administration of laws than depending on politically volatile federal agencies. The data from the current moment provides powerful support for that thesis.

Federal enforcement is binary: it exists or it doesn’t, depending on who holds the White House and what ideology they bring to regulatory appointments. The CFPB went from aggressive enforcement under Chopra to functional elimination under Vought in a matter of weeks. An entire enforcement infrastructure—cases, rules, guidance, complaint resolution—was destroyed in a single administration. This volatility is a feature of centralized enforcement, not a bug. It is inherent in any system that concentrates enforcement authority in institutions controlled by political appointees.

Decentralized enforcement through private attorney general mechanisms does not have this vulnerability. It does not depend on any single institution or appointment. It operates through thousands of individual judges making independent decisions to enforce their own rights under statutes that exist regardless of who occupies the executive branch. An administration can fire a CFPB director. It cannot fire every person who decides to check their billing statement against the statutory requirements.

AI amplifies this resilience by orders of magnitude. The number of potential judges is no longer limited to the small cadre of consumers who happen to have legal knowledge or access to attorneys. It is, in principle, every person with a smartphone—which is to say, nearly every adult in the United States.

The obligations of discovery

With the power of judgment comes obligation. The judge who identifies systemic noncompliance possesses information that is material to the public interest. The private attorney general doctrine is not a license for extortion or bad-faith litigation—it is a framework for vindicating public rights through legitimate legal channels. The judge’s credibility depends on accuracy, documentation, and transparency.

This means filing under penalty of perjury when the mechanism exists to do so. It means disclosing your own interests and potential conflicts. It means retaining original documents and making them available to regulatory authorities. It means directing your findings to the appropriate agencies even when those agencies have been weakened, because the filing creates a record that imposes institutional obligations regardless of the current enforcement climate. It means being willing to be examined and scrutinized, because the judgment’s strength derives from the evidence, not the judge’s authority.

The judge is not above the system. The judge is the system—the mechanism the legislature designed to ensure that public laws are enforced when government institutions fail. That design carries responsibilities, and the AI tools that make discovery possible do not discharge those responsibilities. They make fulfilling them easier.

What comes next

The United States is running a live experiment in what happens when you remove the cop from the beat. The CFPB has been reduced to an agency on paper only. The FTC’s independence faces an existential threat at the Supreme Court. The private attorney general—armed with FDCPA statutory damages, state UDAP claims, self-executing void-contract statutes, and AI-powered compliance analysis available for twenty dollars a month—is being asked to shoulder a burden that federal agencies with billion-dollar budgets once carried.

The question is not whether private enforcement will grow. It already is, measurably and dramatically. The question is not whether AI will transform the judge’s capabilities. It already has. The question is whether a system that depends on individual citizens to detect violations, document them, and deliver judgments can generate justice at scale.

The data from this moment suggests the answer is yes—with caveats. Individual victories are possible. Systemic discovery is possible. Template-level noncompliance can be identified from a single billing statement. The statutory framework to convert that discovery into judgment is already built. The tools to enable that discovery are already in the hands of the people who need them most.

But private enforcement, however resilient, is not sufficient by itself. It cannot write binding rules. It cannot conduct supervisory examinations. It cannot subpoena internal documents without litigation. It cannot coordinate a national response to systemic noncompliance across industries that span all fifty states. The judge is the last line of defense. The judge should not have to be the only one.

The judge is the last line of defense. The judge should not have to be the only one.

Epilogue: Enough

I did not set out to become a judge. I did not set out to file regulatory complaints or draft demand letters or research the fee-shifting doctrines that make consumer enforcement economically viable. I set out to understand why my medical bills looked wrong. The answer turned out to be that they were wrong—not just in amount or in coding, but in their fundamental legal validity. The instruments generating my debt were noncompliant with the statutes governing them, and the debt they purported to create did not legally exist.

One hundred million Americans are carrying medical debt right now. A significant percentage of them have looked at a billing statement that didn’t make sense, felt the weight of it, and paid anyway. They paid because the system is designed to make payment the path of least resistance and questioning the path of maximum friction. The disclosures that would tell them they have the right to dispute are missing from the templates. The information that would let them evaluate the bill’s accuracy is withheld. The institutional apparatus that was supposed to protect them has been dismantled.

The private attorney general doctrine exists because the founders of this legal tradition understood that government enforcement would sometimes fail—that agencies would be captured, defunded, or ideologically redirected. They built a backup system: empower the citizen, arm the judge, let private litigation carry the weight when public institutions cannot. That backup system is now the primary system, and the tools that make it operational are more powerful and more accessible than at any point in American history.

I asked an AI to check my billing statement. The answer came back in seconds. And then I did what the system was designed to let me do: I examined the evidence. I applied the law. I delivered the verdict.

I am the law. Because enough is enough.

I need to tell you something, and I need you to really hear it.

Not the way you hear advice from well-meaning strangers on the internet. Not the way you half-listen to a podcast while scrolling through your phone. I need you to hear this the way you hear a truth you’ve been waiting your whole life for someone to say out loud.

So do me a favor. Stop whatever else you’re doing. Put both feet on the floor if you can. Take one breath—a real one, not the shallow little survival breaths you’ve trained yourself to take when your body feels like enemy territory. Breathe all the way down to your belly. Hold it. And let it go.

Good. Now stay with me.

When you were born, someone looked at your body and made a decision about who you were. They wrote it down on a piece of paper. They told your parents. Your parents told the world. And from that moment forward, every institution, every relationship, every mirror you ever looked into was calibrated to reflect that decision back at you.

The clothes you were handed. The name you were given. The way people touched you, talked to you, expected you to move through the world. All of it was built on a single glance between your legs in the first seconds of your life.

Here’s what no one told you: that glance was not a discovery. It was an assignment. And assignments can be wrong.

You’ve always known this. Maybe not in words—maybe it was just a feeling, a low hum of wrongness that lived in your body like background radiation. Maybe it was the way you flinched getting dressed in the morning, or the strange numbness you felt in moments when everyone else seemed comfortable in their skin. Maybe it showed up as a question you couldn’t quite form, a sentence that started with “What if I—” and never got to finish because the world had already decided the answer for you.

That hum? That flinch? That unfinished question?

Those aren’t symptoms of something broken inside you. They’re you, trying to get your own attention. They’re the signal that the map you were given—the one that says this body means this identity means this life—doesn’t match the territory of who you actually are.

I know this because I lived it for decades. And I know this because hundreds of people have told me their version of the same story, in DMs at 2 a.m., in comments on my posts, in emails that start with “I’ve never told anyone this, but...”

You are not alone in this. You never were.

The Lie That Holds Everything in Place

There is a single, foundational lie at the center of how most of the world understands gender, and it goes like this: your genitals are your gender.

It sounds so obvious when you say it plainly that most people never think to question it. Of course your body determines who you are. Of course anatomy is destiny. Of course the parts you were born with are the final word on the subject.

Except they’re not. They never have been. And the evidence for this isn’t hidden—it’s everywhere, in every culture, across every century of recorded human history. Indigenous communities across North America have recognized Two-Spirit people for generations. South Asian cultures have acknowledged hijra as a distinct gender category for thousands of years. Samoan fa’afafine occupy a social role that predates European contact by centuries.

The rigid, binary, genitals-equal-gender framework isn’t a universal truth. It’s a specific cultural invention—and a relatively recent one at that.

But here’s what makes the lie so sticky, so hard to shake loose even when you know better: it doesn’t just live in laws and policies and medical forms. It lives in your nervous system. It lives in the micro-flinch you feel when someone genders you based on assumptions about your body. It lives in the dissociation you slip into during intimacy because being present in your own skin feels unbearable. It lives in the exhausting, never-ending mental energy you spend just managing a body that doesn’t feel like yours.

That’s what genital dysphoria actually is. Not a delusion. Not confusion. Not a phase. It’s the lived experience of carrying a body that was mapped to an identity you were never meant to inhabit. It’s the bone-deep knowledge that the parts you were given don’t match the person you are.

And if you’ve felt it, you don’t need me to explain it to you. You just need someone to say: I see you. I believe you. And there is a way through this.

What Nobody Tells You About Bodies

Here’s where I need to get real with you, because this is the part that changed everything for me, and it might change everything for you too.

Your body is not a fixed thing. It is not a destiny. It is not a verdict handed down by biology that you are powerless to appeal.

I’m a nonbinary person who underwent genital nullification—a procedure that most people have never heard of, that removes external genitalia to create a smooth, neutral surface. I also take testosterone, because that’s what aligns my mind, my emotions, and my sense of self with who I am. These choices might seem contradictory to someone still operating under the old map. Testosterone and nullification? Masculine presentation and no genitals?

But they’re not contradictory at all. They’re precise. Each decision was a calibration, a fine-tuning of the instrument of my body until it finally played the right note. And when it did—when I woke up from surgery and looked down and saw a body that matched my inner landscape for the first time—the silence hit me in the chest. Not empty silence. Full silence. The kind of silence that comes when a noise you’ve been hearing your entire life suddenly stops, and you realize you’d forgotten what quiet sounded like.

That’s what congruence feels like. And you deserve to feel it too.

But here’s what nobody tells you: there is no single path to that feeling. The world wants to give you a script—first you change your name, then you start hormones, then you get “the surgery” (as if there’s only one), and then you’re done. Transition complete. Credits roll.

Real life doesn’t work like that. Real bodies don’t work like that. The landscape of what’s actually possible is so much vaster and more creative than the standard narrative suggests.

There are men who have had vaginoplasty—men who continue taking testosterone, who live and present as men, whose manhood is not one bit diminished by having a vagina. There are women who have chosen vulvoplasty, claiming their femininity on their own terms, in their own bodies. There are people like me in the nullo community, some of us masculine, some feminine, some beautifully androgynous, all of us united by the choice to create a body that reflects our truest selves rather than anyone else’s expectations.

And then there are the pioneers—people working with skilled surgeons to create configurations that are entirely unique to them. People with both a phallus and a vagina. People who have modified specific aspects of their anatomy while keeping others exactly as they are. People who listened to their dysphoria with such precision that they could target the exact source of their discomfort without sacrificing other parts of their identity that felt true.

These aren’t stories of confusion. They’re stories of extraordinary clarity.

And here’s what I want you to sit with for a moment: every one of these people, at some point, was exactly where you are right now. They were the person lying awake at 3 a.m., wondering if what they felt was real. They were the person Googling terms they’d never heard spoken aloud, afraid of what they’d find and more afraid of not looking. They were the person who thought they were the only one in the world who felt this specific, precise way about their body—until they found a community, a word, a mirror, and realized they weren’t.

The distance between where you are and where they are is not a canyon. It’s a path. And people have walked it before you, and they’ve left markers along the way. You are not inventing this journey from scratch. You are joining a lineage of people who chose themselves, and that lineage is older and wider and more beautiful than you might imagine.

The Weight of Waiting

There’s something I want to name that doesn’t get talked about enough, and it’s this: the waiting is its own kind of suffering.

I don’t just mean waiting for a surgery date, though that waiting is real and can be agonizing. I mean the years—sometimes decades—of waiting before you even give yourself permission to want something different. The years of thinking maybe it’s not that bad. The years of bargaining with your own body, trying to find a way to make peace with something that was never going to fit. The years of watching other people live comfortably in their skin and wondering what that must feel like, the way a person born blind might wonder about color.

That waiting has a weight to it. It presses down on you. It shapes the way you move through the world—carefully, guardedly, always a little bit dissociated, always spending some portion of your mental energy on the project of managing rather than living. You develop coping mechanisms that look like personality traits. You learn to avoid mirrors, or to only look at certain parts of yourself. You learn to change clothes in the dark. You learn to leave your body during intimacy, to go somewhere else in your mind while your physical self does what’s expected of it. You learn to perform comfort you don’t feel, ease you don’t have, normalcy that costs you everything.

And the world around you has no idea. Because you’ve gotten so good at it. Because you’ve been doing it for so long that you’ve almost convinced yourself it’s fine. Almost.

But almost is not enough. Almost is its own kind of prison.

I want to tell you that the waiting can end. Not all at once, and not without effort, and not without navigating a system that wasn’t built for people like us. But it can end. The first step isn’t a surgery date or a hormone prescription. The first step is this: admitting to yourself that you want something different. That’s it. That’s the beginning. Everything else flows from that one act of honesty.

And if you’re reading this article, you may have already taken that step. You may have taken it years ago. In which case, what you need now isn’t permission to want—it’s a roadmap for getting there. We’ll come back to that.

The Permission You’ve Been Waiting For

I want to address something directly, because I think it might be the reason you’re still reading.

You’re waiting for permission.

Not from me, specifically—though if it helps, you have it. You’re waiting for permission from somewhere, from someone, to trust what you already know about yourself. You’re waiting for an authority figure, a medical professional, a parent, a partner, a community to look at you and say: Yes. What you feel is real. What you want is valid. You are allowed to do this.

I understand that wait. I lived in it for years. I performed the version of myself that I thought would earn me access to care, to acceptance, to basic respect. I learned the “right” answers to give therapists. I learned to frame my experience in terms that gatekeepers would recognize. I learned to shrink my actual, complicated, nonbinary, nullo truth down into a shape that fit inside someone else’s checkbox.

And let me tell you what I learned from all that performing: it almost killed me. Not metaphorically. The gap between who I was pretending to be and who I actually was became a wound that got wider every year. The weight of that pretense, that constant self-betrayal, was heavier than any dysphoria I’d ever felt.

So here is what I need to say to you, as clearly and as directly as I can:

You do not need anyone’s permission to be who you are.

You do not need to convince a panel of strangers that your pain is real enough. You do not need to perform a specific version of gender to be “trans enough” for care. You do not need to follow someone else’s script for what transition looks like. You do not need to justify your desire for a body that feels like home to anyone—not a doctor, not a therapist, not your family, not the internet.

Your internal sense of peace is the only metric that matters. Your comfort is not a frivolous request. It is a medical necessity. Your peace is reason enough.

The era of begging for belief is ending, because we are ending it. Patient by patient, story by story, we are dismantling the system that required us to prove our own reality to strangers before we could access the care we needed. We are building a world where informed consent is the standard—where the only expert required to validate your identity is you.

The Words That Set You Free

Part of claiming your body is claiming the language you use to talk about it. This matters more than you might think.

For most of us, the only words we were ever given for our bodies were clinical ones. Medical terms that describe procedures, not people. Words that make you feel like a specimen on a table rather than a human being finding their way home. Those words have their place—they’re useful for communicating with surgeons, for reading research papers, for understanding what’s medically possible. But they are not your words. They are the operating room’s words.

So we made our own.

In communities like mine, language has evolved from the ground up—in whispered conversations, in late-night forum posts, in the shared vocabulary of people who finally found each other and realized they weren’t alone. Terms like “nullo” didn’t come from a textbook. They came from us. And finding a word that fits—a word that describes your experience with accuracy and warmth and sometimes defiance—can feel like coming home in miniature. It’s a small act of reclamation that carries enormous power.

You have the right to name yourself. You have the right to name your body, your parts, your experience, in whatever words feel true. Medical terms, community terms, words you invented in the shower at 3 a.m. that no one else has ever heard—all of it is valid. All of it is yours.

Try words on. See how they feel. Discard the ones that don’t fit. Keep the ones that make you feel seen. And know that this is an ongoing process—language is alive, and your relationship with it will grow and shift just like your relationship with your body. That’s not inconsistency. That’s evolution.

The World Right Now (And Why This Matters More Than Ever)

I’d be lying to you if I didn’t acknowledge the moment we’re living in.

As I write this, gender affirming care is under legislative attack in ways that would have seemed unthinkable a decade ago. Legislators are introducing bills in state after state to restrict access to hormones, ban surgeries, force detransition, and criminalize the very care that saved my life and the lives of countless people I love. Insurance companies are finding new ways to deny coverage. Politicians who have never met a trans person are making decisions about our bodies with the confidence of people who have never once questioned their own.

It would be easy to feel hopeless in a moment like this. It would be understandable to retreat, to go quiet, to put your own journey on hold and wait for a better political climate.

I’m asking you not to do that.

Not because the threats aren’t real—they are, and they require our collective resistance, our votes, our voices, our refusal to be erased. But because your life is happening now. Your body is yours now. The dysphoria you’re carrying doesn’t pause for election cycles. Your need for care doesn’t wait for a friendlier legislature.

The people who want to restrict our access to care are counting on us to internalize their message—to believe, on some level, that we don’t deserve what we’re asking for. That our needs are frivolous, or dangerous, or too complicated to be taken seriously. Every time you advocate for yourself, every time you walk into a doctor’s office and say this is what I need, every time you refuse to shrink yourself to fit someone else’s comfort level, you are proving them wrong. You are living proof that this care matters, that these choices are ours to make, and that no law or policy can override the fundamental human right to be at home in your own body.

This is why the practical knowledge matters so much right now. Knowing your options, knowing your rights, knowing how to navigate the system—this isn’t just helpful information. In this political moment, it’s armor. It’s power. It’s survival.

The Joy on the Other Side

I’ve spent a lot of this article talking about pain, and that’s necessary, because pain is real and it deserves to be acknowledged. But I would be doing you a disservice if I didn’t also talk about joy.

Because here’s the thing about aligning your body with your inner self: it’s not just the absence of suffering. It’s the arrival of something you stopped believing was possible. It’s a quiet, bone-deep peace that settles into you like warmth after a long winter. It’s the ability to catch your reflection and feel—maybe for the first time in your life—recognition. Not the anxious, dissociated scanning you’ve been doing for years, checking for what’s wrong, cataloging what doesn’t match. Just... recognition. Oh. There I am.

I remember the first time I looked in the mirror after my nullification surgery and didn’t flinch. I just looked. And the person looking back was, finally, undeniably, me. Not a costume. Not a compromise. Not a performance for someone else’s benefit. Me.

That moment is available to you. Not necessarily through the same path I took—your body, your journey, your map. But the destination? The feeling of being at home in your own skin? That is your birthright, and no one has the right to keep it from you.

The energy you’ve been spending on managing dysphoria—the constant vigilance, the dissociation, the negotiation with a body that doesn’t feel like yours—that energy gets freed up when congruence arrives. And what you do with that energy is extraordinary. You invest it in your relationships, your passions, your work, your play. You stop surviving and start living. It changes everything.

Every step you take toward that feeling—whether it’s as small as buying a piece of clothing that makes you feel like yourself, or as significant as walking into a surgeon’s office—is an act of defiance. In a world that demands conformity, choosing yourself is the bravest thing you can do.

One more thing: the joy doesn’t erase the grief. You will mourn the years lost to dysphoria, the life you might have lived with earlier access to care, the relationships that didn’t survive your authenticity, the ease that cisgender people take for granted.

But the grief and the joy coexist. You look back and feel sadness and look at your present self and feel a gratitude so fierce it stings. These aren’t contradictions. They’re the texture of a life lived honestly.

That honest life, the messy one where you stopped performing and started being, is richer than what came before. Even on the hard days. Because the person experiencing all of it is finally you.

Why I Wrote the Book

Last year, I published my Substack chapter “Your Body, Your Rules” and something happened that I didn’t expect. My inbox exploded with dozens of messages from people all over the world—trans men, trans women, nonbinary folks, people who had never told anyone what they were feeling, people who had been searching for years for someone who understood.

The messages all had something in common. They said, essentially: Thank you. I’ve never seen my experience reflected back to me before. But I need more. I need the practical stuff. I need to know HOW.

How do I find a surgeon who will actually listen to me? How do I pay for this? How do I navigate insurance, and legal documents, and the medical system that still doesn’t understand people like us? How do I prepare for surgery, and recover from it, and rebuild my relationship with my body afterward? How do I do this if I’m neurodivergent, or disabled, or living in a country where this care barely exists? How do I do this if I’m terrified?

One Substack article couldn’t answer those questions. A whole book could.

So I wrote Becoming Yourself: The Genital Dysphoria Survival Guide.

It is the book I wish had existed when I was standing at the beginning of my own journey, overwhelmed and alone and not knowing where to start. It covers everything—and I mean everything:

The emotional and intellectual foundation: Understanding dysphoria not as pathology but as signal. Separating your body parts from your identity. Debunking the myths that keep us small. Reimagining what embodiment can look like when you’re the one drawing the map.

Every major surgical option in detail: Nulloplasty, vaginoplasty, vulvoplasty, orchiectomy, penectomy, hysterectomy, metoidioplasty, phalloplasty, top surgery, facial surgery, voice surgery—each one its own chapter, written with both clinical precision and the warmth of someone who has been through it.

The practical survival guide: How to find and vet surgeons. How to navigate insurance and financing in the U.S. How to deal with the legal and administrative maze. Fertility preservation. Recovery essentials—both for you and for the person taking care of you. What to expect from your sexuality and intimacy in a new body. How to document your surgical journey. How to thrive long-term.

Specialized chapters for those who need them: A full chapter on neurodivergence, disability, and accessibility in gender affirming care. A chapter on managing dysphoria and expectations while you wait. A chapter on navigating international care for readers outside the U.S.

Country-by-country guides for accessing gender affirming surgery in Canada, Mexico, the UK, France, Germany, the Netherlands, Spain, Italy, Turkey, Thailand, Japan, South Korea, Australia, and New Zealand—because this community is global, and so is this book.

And a chapter on what’s happening right now: the political landscape, the legal threats, the ways our healthcare is under siege—and what we can do about it.

This isn’t a pamphlet. It’s not a blog post series stitched together. It is the most complete resource on genital dysphoria and gender affirming surgery that exists, and I wrote every word of it with you in mind.

This Is Your Invitation

If the words in this article resonated with you — if something loosened while you were reading — then I want you to know that everything I wrote here came from a much larger project.

Last year, when my inbox filled with messages asking how — how do I find a surgeon, how do I pay for this, how do I survive the system — I realized one article wasn’t enough. So I wrote the book I wish had existed when I was starting this journey alone.

Becoming Yourself: The Genital Dysphoria Survival Guide is available now. You can get your copy here:

Buy the book now on Gumroad!

I wrote it because the version of me who was six years old, drawing stick figures in kindergarten, trying to tell someone that something was wrong — that kid deserved a book like this. And so do you.

That silence I mentioned — googling “nullification surgery” at midnight and finding nothing, no guides, no stories, no one saying I did this and here’s what happened — that silence is what this book exists to break. Not with abstract philosophy, but with the specific, practical knowledge you need to get from where you are to where you want to be.

If you’re not ready yet, that’s fine. Bookmark it. Come back when you are.

You are becoming yourself. And I wrote this so you wouldn’t have to do it alone.

Three Questions

Before you close this page:

1. What would your body feel like if you stopped listening to everyone else’s opinion about it?

2. What’s one thing you could do this week to honor what you already know about yourself?

3. What changes the moment you stop waiting for permission?

You don’t owe anyone the answers. But don’t let the questions go.

Axl Ibiza is the author of Becoming Yourself: The Genital Dysphoria Survival Guide, the most complete resource on genital dysphoria and gender affirming surgery ever written. The book is available now on Gumroad. Follow Axl on Substack for ongoing community, resources, and support.

The difference between professionals who achieve mediocre results from AI and those who achieve transformative results isn’t the model they use—GPT-4 versus Claude—or their prompt engineering skills.

The real difference is context: the depth of understanding, clarity of organization, and accessibility of knowledge you’ve systematically accumulated in your notes. Put simply: AI output quality depends on context quality. Whether it’s your codebase, your notes, or any other data, the principle holds.

I’m also speaking to people who don’t like AI or are actively against it. This article is still for you. This isn’t just about AI. Your personal knowledge base—in a system like Obsidian or Notion—is your bedrock source of truth, your archive, your library for tracking the information, facts, sources, and details that matter most in your life. Creating rich, detailed notes is the only way to make sense of the complexity we live in. The world has too much noise. We all need systems to filter and track it.

I started my personal knowledge management system at five years old with my first diary, which I still have. Born in the late 1980s, this meant digitizing everything from paper to PDF. Thankfully, I finished most of that work five years ago—no more paper sitting around my house. It’s all on my NAS server now. The real question is: how do you create a clean, unified context for your models and agents to truly know who you are? Today, for example, I’ve been working on migrating my recipe collection from my years as a chef.

This is the critical inflection point in the evolution of knowledge work. If you haven’t started treating your notes as critical infrastructure—as a foundational asset deserving the same attention as any other professional tool—you’re about to feel the consequences.

This week, I revisited one of my dream projects—an ambitious idea that always seemed achievable but required too much grunt work to be realistic. With AI, it’s now actually possible. From my decade as a chef, I have a massive recipe collection in every format imaginable: handwritten notebooks, cookbooks, documents from restaurants I’ve worked in, James Beard kitchens. It’s a rich tapestry of food knowledge I’ve always wanted to consolidate into a single source.

I’ve always been a fan of the Paprika app, but the most important part of this workflow is converting all these sources into clean YAML data that any software can read. You could use Notion or Obsidian instead. From personal experience cooking professionally and relying on my phone for recipes, I trust Paprika because every recipe comes from pristine sources that work. None of it is hallucinated or made up.

I don’t need hallucinated AI recipes when I’m working professionally. I would never want AI writing recipes from scratch—that’s not what I’m asking. I’m asking it to convert existing recipes into clean data. Every chef knows the power of carrying a little black book—in your pocket, pants, or shirt—where you write everything down so you don’t forget it. It’s a habit we could all use, whether analog or digital. When this data migration project of mine is done, every single recipe I’ve ever worked with is going to be merged into a single app on my phone, written in YAML, Markdown’s cousin. In other words, my entire database of recipes that I trust will be instantly available to me at all times.

Even when I was cooking professionally, data was my superpower. I started digitizing my recipes in culinary school, and it’s the reason I’m so effective as a chef—I always have the best information at my fingertips. The principles in this article apply across all skill sets. It doesn’t matter what kind of work you do. You need notes. You need historical context. Otherwise, how do you improve over time? If you don’t know what worked, what didn’t, which data is good, which is bad, which sources you can trust, and which you can’t—you’re flying blind. The scientific method is built on data.

The Tale of Two Knowledge Workers

Let me paint you a picture of where we are in late February 2026, because the differences matter.

Meet Jordan, a content creator who’s been building an audience for two years. Jordan uses every AI tool available—Claude, GPT-4, Gemini. Every morning, Jordan spends two hours in chat interfaces asking about content strategy, article outlines, and research synthesis. The answers are coherent and mostly correct, but they’re generic—indistinguishable from what any creator might receive.

The problem reveals itself in the details. Jordan asks about “newsletter topics,” and the AI suggests standard creator advice without knowing the audience’s specific interests. Jordan asks about “content structure,” and the AI provides generic best practices without understanding what the readers actually need. Jordan asks about “competitive positioning,” and the AI mentions generic competitors without recognizing what makes Jordan’s work unique.

Jordan is busier than ever, but outputs don’t feel more strategic. They feel more generic. Cognitive energy gets burned validating and correcting AI suggestions instead of building on them.

⚡Now let me tell you about my own system, because this is where things get real.

I’ve spent the last year building something fundamentally different: not just using tools well, but architecting an entire ecosystem. My knowledge base isn’t trapped in a single application—it’s a web of interconnected data that continuously interfaces across platforms. This is the critical shift: it’s no longer about mastering one piece of software. It’s about creating an ecosystem of knowledge and agents that continuously compound and work on your behalf.

I don’t just use Notion. I don’t just use Claude. I don’t just use Obsidian or any single tool. My system is an architecture: automated pipelines capture newsletters as they arrive, AI systems extract content and convert it to clean Markdown, databases structure everything with metadata and relationships, backup systems run hourly snapshots, and agents query this entire ecosystem regardless of which interface I’m using.

On any given day, I can use Claude on my computer, Notion AI in my workspace, GPT-4 through an API, or any other provider and get similarly excellent results—not because the models are the same, but because they’re all grounded in my data. My 1,247 processed newsletters. My 34 research papers from this month alone. My years of cross-referenced community insights. My structured documentation of patterns, failures, and successes.

When I need to write an article, I don’t start from scratch. My ecosystem has already done the groundwork. My AI agents query my entire knowledge base: every newsletter summary, every research finding, every community discussion, every resource I’ve cataloged over years of deep work. The AI returns a first draft already grounded in my actual expertise—referencing specific sources my system archived, connecting insights across research streams, flagging where new information contradicts older assumptions.

I don’t spend hours editing and fact-checking. The bulk of the work is already done, grounded in real, accumulated domain expertise from my knowledge base. Every recommendation connects to evidence. Every argument links to specific sources from my automatically-archived collections.

Jordan spends 40 hours a month on research and synthesis work and still feels like something is missing.

I spend four hours a month and have more confidence in both the output and the reasoning. The difference isn’t the tool. The difference isn’t prompt engineering skill. The difference is infrastructure.

I built automated pipelines that continuously feed my knowledge base. I built AI systems that classify and extract insights with 94.2% accuracy. I built a structured ecosystem that serves as a queryable second brain—accessible through any AI interface, any platform, any tool that can connect to my data. Jordan hasn’t.

This difference becomes visible in output quality and sustainability. When I wrote Becoming Yourself: The Complete Guide to Gender-Affirming Surgery—a 40-chapter comprehensive book—I had years of structured knowledge to draw from. The book didn’t require starting from zero; it required synthesizing what my ecosystem had already captured and organized.

Jordan, meanwhile, starts every major project from scratch. Rediscovering insights encountered months ago but never systematically captured. Re-researching topics already covered. Unable to easily reference past work because it’s scattered across disconnected tools, email threads, and half-remembered conversations never documented in a structured, AI-queryable way.

This gap is about to become a chasm—and then an unbridgeable gulf that fundamentally restructures who has power in knowledge work.

Within 18 months, the professionals who’ve spent years building structured, AI-queryable knowledge ecosystems will operate at a velocity that makes their peers look functionally obsolete. They won’t just be “more productive.” They’ll be playing a different game entirely.

Let me be clear about what I mean by “ecosystem.” I’m not perfecting the use of a single system. I’m not trying to become a Notion expert or an Obsidian power user or a master of any one tool.

I’m chaining all my data silos together into interconnected webs that talk to each other on their own.

My newsletters don’t just sit in my inbox. They’re automatically processed, converted to Markdown, tagged with metadata, and stored in a queryable database. My research papers aren’t trapped in PDFs. They’re extracted, summarized, and cross-referenced with related notes. My meeting transcripts aren’t isolated files. They’re parsed, linked to relevant projects, and connected to action items that surface when I need them.

This is the fundamental shift: from using tools to architecting ecosystems. From manually moving data between systems to building automated pipelines that do it for me. From searching for information to having information surface when relevant.

My Obsidian vault talks to my Notion workspace. My email automation feeds both. My backup systems mirror everything hourly. My AI agents can query across all of it, regardless of which interface I’m using. When I’m in Claude, it sees my vault. When I’m in Notion, it sees my databases. When I’m using an API, it sees everything.

This isn’t about mastering one piece of software. This is about building infrastructure that compounds—where every new piece of information automatically connects to everything else, where every insight gets tagged and linked and made retrievable, where the system itself maintains connections I’d never maintain manually.

Jordan will spend 40 hours researching a strategic decision, synthesizing information from scratch, validating sources manually, and still second-guessing whether everything relevant was found. I’ll spend four hours querying my knowledge ecosystem—1,247 automatically processed newsletters, 34 research papers from this month, years of cross-referenced insights—and produce recommendations grounded in verifiable evidence that my system has already organized, tagged, and connected.

The economic implications are stark. Every strategic decision Jordan makes costs 10× more in labor hours. Every article requires 3× more editing because the AI had no context. Every pitch feels generic because the system contains no accumulated intelligence. Jordan isn’t lazy or unintelligent. Jordan is simply competing without the infrastructure that makes modern knowledge work sustainable.

Here’s what happens next: Organizations hire people like me. Not because we’re smarter, but because we can deliver the same quality output in one-tenth the time, with verifiable sourcing, with clear reasoning chains, with confidence in the recommendations. Our knowledge ecosystems are our competitive moats. They’re worth tens of thousands of dollars in saved labor hours annually. They compound every month.

The professionals who don’t see this coming will wake up in 2027 wondering why they’re losing clients to competitors who somehow “just know more.” The answer won’t be that those competitors are smarter. It will be that those competitors built infrastructure in 2024–2026—interconnected ecosystems of data and agents—while everyone else was still treating notes as throwaway artifacts and hoping a single perfect tool would solve everything.

This isn’t about productivity hacks. This isn’t about mastering one piece of software. This is about who controls valuable context in an AI-mediated economy through well-architected knowledge ecosystems. Your structured, interconnected knowledge base is becoming your most valuable professional asset. The gap between people who understand this and people who don’t is about to become a chasm. And that chasm is about to become a permanent structural divide in who has power, who gets hired, and who sets the terms.

How We Got Here: The Evolution of Knowledge Management

To understand why this moment matters so much, we need to look back at how knowledge management has changed, because that history explains everything about where we are now.

For most of the twentieth century, knowledge work meant filing cabinets. Folders. Paper. If you needed to remember something, you either kept it in your head or you filed it away and hoped you could find it later. The constraint was storage. If you had a good filing system—and most people didn’t—you could accumulate maybe a hundred important documents. Everything else was lost.

Then computers arrived, and the constraint shifted. Storage became cheap. The new problem was organization. In the 1990s and 2000s, we tried to solve this with rigid hierarchies. Folders nested inside folders nested inside folders. You had a “Projects” folder, and inside that was “2024,” and inside that was “Project X.” By 2015, most knowledge workers had forgotten what was in half their folders because the rigid hierarchy didn’t match how their brain actually worked.

Enter Evernote in 2000. For the first time, you could dump everything into a single system, tag it loosely, and search for it later. The constraint shifted again. Now it wasn’t storage or organization—it was capture rate. Evernote was so frictionless that you could actually maintain it. You could save articles, voice memos, photos, all mixed together. For a moment, it felt like the knowledge management problem was solved. Thousands of people built their lives around Evernote. It became ubiquitous.

But then two things happened simultaneously. First, Evernote’s growth stalled. The company failed to innovate. By 2020, people were looking for alternatives. Second, and more importantly, artificial intelligence suddenly got useful.

When ChatGPT launched in November 2022, nobody was thinking about knowledge management. It was just a neat chatbot. But within a year, something became obvious to a small group of early adopters: when you gave an AI system access to your files, your notes, your documents—when you used Retrieval-Augmented Generation (RAG)—the AI could become genuinely useful for your specific problems, not just generic questions.

People started asking: what if I could feed my entire vault of notes into an AI and ask it to synthesize? What if I could ask my AI system to write a decision document, and it already had access to all the research and context I’d accumulated?

The people who tried this realized something: it worked. But only if your notes were structured a certain way.

Evernote’s proprietary format was a black box. Notion’s database structure was powerful but locked in. Google Docs’ web-based architecture wasn’t designed for bulk AI ingestion. And suddenly, an old format became wildly relevant: Markdown.

This is the moment we’re in now. And it’s fundamentally different from every previous era.

The constraint is no longer storage or organization or searchability. The constraint is now structure and accessibility to AI systems. Your notes need to be in a format that AI can read, parse, and understand at scale. They need to have metadata that AI can query. They need to have links that AI can follow to understand relationships between ideas. They need to be atomic—one idea per note—so that AI can retrieve and recombine them intelligently.

Markdown is the format built for this world. Plain text, universal compatibility, machine-readable structure, human-readable content. It’s been around since 2004, but it’s only now, in 2026, that it’s become strategically important.

And Obsidian—a local-first, Markdown-based note system—has emerged as the tool that takes this seriously. Not the only tool, but the one that’s most aligned with this future. Roam Research was first, but Obsidian has become the de facto standard because it respects a principle that matters more every month: your notes should be yours, stored on your computer, in an open format, portable to anywhere.

The professionals who understand this shift are already building their vaults. The ones who don’t will wake up in six months and realize they’re competing against peers who have a two-year head start on structured knowledge.

The Economics of Context: Why This Matters Beyond Productivity

This isn’t just about productivity. It’s about asymmetric advantage.

Major technology companies have spent billions on a key insight: properly structured proprietary data is worth millions. That data fine-tunes AI models to outperform generic alternatives.

Consider Google. They had raw internet data, but something more valuable: click logs, search queries, and user behavior patterns from billions of searches. They used that proprietary data to build models specifically good at ranking web pages. That competitive advantage made Google the dominant search engine—worth hundreds of billions of dollars.

Anthropic, OpenAI, and other AI labs have published extensively on this. A generic model trained on internet data provides baseline capability. But a model fine-tuned on your company’s proprietary data—code repositories, documentation, customer interactions, internal research—can outperform generic models by orders of magnitude for your specific use case.

This explains why companies spend millions on Retrieval-Augmented Generation (RAG) infrastructure. Why Microsoft spent billions integrating AI into enterprise products. Why Salesforce built Einstein, integrating AI with customer data.

Companies protect this data fiercely. They encrypt it, version control it, and invest in retrieval systems. They build RAG pipelines that give AI agents the right context at the right time. They hire data engineers and ML experts to ensure proprietary data is structured for AI use.

Why? Because the data generates productivity gains. And productivity gains compound into competitive advantages.

Here’s a concrete example. A software company with 500 engineers built a RAG system for their codebase and documentation. They trained an AI agent to answer questions about architecture decisions, implementation patterns, and best practices. According to internal analysis, this reduced code review time by 30%, onboarding time by 40%, and architectural mistakes by 25%. The productivity gains were worth millions in annual engineering time.

For individuals, your personal knowledge base works the same way.

Your notes are your proprietary data. The insights you’ve accumulated through work, research, failures, and pattern recognition aren’t available on the internet. They’re not in any training dataset. They’re unique to you—valuable specifically because they’re yours.

When you give an AI agent access to this knowledge, you create something without equivalent: an AI system optimized for your problems, your context, your decision-making. It’s not better than generic AI at abstract tasks. But it’s exponentially better at helping you make decisions in your specific domain.

That advantage translates to real money across every knowledge profession. For managers, it’s the difference between decisions that stick and decisions requiring rework, with every hour spent reworking a bad decision representing an hour not spent building. For writers, it’s the difference between researching a topic from scratch versus querying years of structured reading notes, article drafts, and research insights—the difference between first drafts needing heavy editing versus light editing, potentially three hours saved per article. For knowledge workers more broadly, it’s the ability to quickly surface relevant quotes, statistics, and sources already collected and verified, rather than re-researching topics covered before. Every piece builds on accumulated insights rather than starting from zero.

For sales professionals, it’s the difference between generic pitches and pitches grounded in customer context—which research shows increases close rates by 15–25%. For consultants, it’s the difference between theoretically sound recommendations and recommendations grounded in accumulated expertise, with every recommendation connecting to evidence and every argument linking to specific sources. For researchers, it’s the difference between starting from scratch and building on ten years of structured research notes queryable in seconds.

The question is: how much is that worth to you? Because the person across the table—the one with two years of structured notes—is capturing that value. You’re not.

And the gap compounds every month you wait.

For Students: The Compounding Advantage You Can’t Afford to Miss

If you’re a student reading this, you have an advantage that I didn’t have when I went through my finance MBA: you know about this system now, before you’ve accumulated years of unstructured notes.

I didn’t start working at Obsidian until January 2023. That’s when I discovered this approach to knowledge management. And that date—January 2023—is basically the starting point of the structured data in my knowledge base. Everything before that? Lost. Scattered across Google Docs, email threads, random notebooks, Evernote dumps that I never organized properly.

Could I migrate all that old material into a structured system? Theoretically, yes. Practically? It would take hundreds of hours to extract insights from old MBA case studies, reformat lecture notes, restructure project reports, and properly tag and link everything. I haven’t done it. Most people don’t. The effort required to retroactively structure years of accumulated knowledge is so high that it almost never happens.

That’s the tragedy. All that learning—thousands of hours of MBA coursework, case studies, group projects, research papers—most of it functionally disappeared because it wasn’t captured in a system that made it retrievable and reusable.

But here’s what happened after I started using this system in early 2023. Within six months, I noticed something strange: I was getting better at synthesizing complex topics than people with far more experience than me. Not because I was smarter. But because I had a system that let me compound insights.

Every article I read got properly structured with metadata. Every client conversation got logged with context. Every project insight got linked to related frameworks. Every decision got documented with rationale and alternatives considered. And within a year, I had a knowledge base that made me more effective than professionals who’d been in the field for a decade but were still operating from memory and scattered notes.

By 2025, the gap had become almost absurd. I could write a strategic analysis in two hours that would have taken me two weeks during my MBA—not because I was working faster, but because I could query years of structured research, pull relevant case studies I’d already analyzed, and synthesize insights I’d already documented and connected. The system did the heavy lifting. I just had to think.

And here’s the uncomfortable truth: grad school started to seem kind of stupid.

Not because the professors weren’t smart. Not because the material wasn’t valuable. But because the entire educational model is built around temporary knowledge transfer. You cram for exams. You write papers. You do group projects. Then you forget 80% of it within six months because you never built a system to retain it.

The implicit assumption in traditional education is that you’ll remember what matters and look up the rest later. But that’s not how knowledge work actually works in 2026. What matters is having a queryable, structured system of everything you’ve learned, connected in ways that let you see patterns and relationships you couldn’t see when you first learned it.

My MBA cost me two years and six figures. If I’d had this system from day one—from the first week of my first semester—I would have captured every case study, every framework, every lecture insight, every group discussion, every paper I wrote, all properly tagged, linked, and structured. By graduation, I would have had a knowledge base worth hundreds of thousands of dollars in accumulated, synthesized, queryable expertise.

Instead, I have some PDFs in a folder somewhere and vague memories of Porter’s Five Forces.

If you’re a student right now, you have a choice I didn’t have. You can start building this system today. Not tomorrow. Not after you graduate. Today.

Here’s what that looks like practically:

Set up Obsidian (or another Markdown-based system) this week. Don’t wait. Don’t overthink it. Download it, create a vault, start capturing. Here’s the truth: the best knowledge management system is the one you’ll actually use. I’ve seen people spend months researching the “perfect” tool, comparing features, reading reviews, watching tutorials—and never actually start building their knowledge base. Meanwhile, someone else downloaded Obsidian on a Tuesday, started capturing notes immediately, and now has six months of structured knowledge while the researcher is still in analysis paralysis. It doesn’t matter if you use Obsidian, Logseq, Roam Research, or even a well-organized folder of Markdown files. What matters is that you start today and build the habit of capturing your learning in a structured, AI-queryable format. You can always migrate or refine your system later. You can’t get back the knowledge you failed to capture while you were still deciding which tool to use.

• Create note templates for the types of content you encounter. Lecture notes. Reading notes. Case study analyses. Project logs. Research papers. Each template should have consistent metadata (date, course, topic, tags, confidence level).

• Make it a daily habit to structure your notes properly. Don’t just copy-paste lecture slides. Synthesize. Add your own commentary. Link concepts to other courses. Tag recurring themes.

• Build a habit of linking related concepts. When you learn about supply chain optimization in operations management, link it to your finance notes on working capital management. When you study game theory in economics, link it to your strategy notes on competitive positioning. These connections are what make the system valuable.

• Document your projects with decision logs. Every group project, every case study analysis, every strategic recommendation you make—document the decision, the alternatives you considered, the rationale, the outcome. This becomes a portfolio of how you think, which is infinitely more valuable than a list of courses you took.

• Review and refine weekly. Spend 30 minutes every Sunday reviewing your notes from the week. Add missing links. Consolidate duplicate ideas. Update your mental models based on new learning.

The effort required is maybe two extra hours per week. That’s it. Two hours per week to structure your learning properly. In exchange, you get a compounding knowledge base that makes you more effective every single month.

By the time you graduate, you’ll have something no other graduate has: a structured, queryable record of everything you learned, connected in ways that reveal insights your professors never explicitly taught.

And here’s the economic reality: employers don’t care about your GPA. They care about whether you can solve problems. And the student who can pull up a case study they analyzed two years ago, synthesize it with three related frameworks, and deliver a recommendation in 30 minutes—that student gets hired. That student gets promoted. That student sets terms.

The student who crammed for exams and forgot everything six months later? That student is competing on credentials, not capability. And credentials are becoming worth less every year.

I started this system in January 2023. It’s now February 2026. That’s just over three years. And the advantage I have over peers who don’t use this system is structural. It’s not that I’m smarter or work harder. It’s that I have infrastructure they don’t have. Every month I use this system, the gap gets wider.

You can start today. The question is: will you?

Because the only way to build this advantage is to start. And the sooner you start, the sooner you begin compounding. There’s no shortcut. There’s no way to catch up later without doing the work. You either build the system now, or you compete without it for the rest of your career.

I didn’t have this advantage during my MBA. You do. Don’t waste it.

What “AI-Ready” Notes Actually Look Like

This is where abstraction becomes concrete. “Structure your notes” is useless advice without specifics.

Don’t let the technical details below intimidate you. I’ll be honest—I was terrible at keeping structured metadata for my notes in Obsidian until two days ago, when I had Claude write code to go through my entire vault and systematically add clean YAML metadata to every note. There’s no need to spend hours of your life cleaning up this data manually. Let AI do it for you.

Start with what most people’s notes look like: unhinged fucking chaos. One note says “Customer feedback—pricing too high.” Another says “pricing thoughts.” A third says “Q3 feedback—need to revisit pricing.” Maybe there’s a Google Doc titled “Product Ideas” with unrelated thoughts mixed in. Maybe a Slack thread with insights buried in a conversation from six months ago that you can’t find anymore.

When an AI tries to make sense of this, it surfaces all three notes and asks you to reconcile them manually. You’ve wasted everyone’s time. The AI has made your job harder, not easier, because now you have to do the synthesis work on top of the AI’s output.

Here’s what AI-ready notes look like. Same information, different structure.

Example 1: Before and After—Customer Feedback

Before (what most people do):

Talked to Sarah at Acme Corp today. She said pricing is confusing.
Other customers have mentioned this too. Need to fix the pricing page.
Maybe also need to revisit our model.

After (AI-ready):

---
type: customer-interview
date: 2026-02-15
customer-name: Sarah Chen
company: Acme Corp
segment: enterprise
product-area: pricing
sentiment: critical
source: sales-call
---

# Customer Interview: Sarah Chen, Acme Corp

## Key Feedback
Sarah identified friction in the pricing model during the onboarding process. She expected a per-user pricing model but found the per-feature model confusing.

## Quote
"I can't tell if we're getting charged for the whole team or per person. It makes it hard to justify the purchase."

## Impact Assessment
This aligns with feedback from 3 other enterprise customers in Q1 2026.

## Next Steps
- Share clarified pricing page with Sarah
- Schedule follow-up call for Feb 28

## Related Notes
- [[pricing-model-feedback]] - consolidated feedback
- [[pricing-competitive-analysis]] - how competitors handle this
- [[pricing-decision-2026-Q2]] - strategic decision framework
- [[onboarding-friction-points]] - related to discovery issue

See the difference? The second version has:

• Metadata (type, date, customer segment, sentiment): Makes it queryable. You can ask “show me all critical sentiment feedback from enterprise customers in Q1” and get an exact answer.

• Structured fields (Key Feedback, Quote, Impact): Makes it parseable by AI. The AI doesn’t have to guess which part is feedback and which part is speculation.

• Backlinks ([[pricing-model-feedback]]): Connects it to your broader knowledge system. When you update the consolidated feedback note, the AI can follow these connections to understand the full scope of the issue.

• Clear source attribution: You know where this came from and when, which matters when you’re synthesizing research six months later.

• Quantified impact: “3 other customers” gives weight to the insight. The AI can search for all notes with similar impact levels.

• Next steps: Transforms a note from a record of what happened to an actionable record of what should happen next.

When an AI reads this note, it understands context. It can group related feedback. It can spot patterns. It can tell you which insights are outliers and which are systemic. It can flag inconsistencies: if you’re hearing that the pricing page is confusing from multiple customers, but your last decision log said the pricing model was clear, the AI can flag that contradiction.

Example 2: Decision Logs

Most people don’t keep decision logs at all. Or they bury them in email threads. Some keep them in Confluence, which locks them in a proprietary system.

AI-ready decision logs look like this:

---
type: decision
date: 2026-02-10
decision-id: PRICING-2026-02
priority: p1
status: approved
owner: alex@company.com
related-area: product-roadmap
impact-area: [engineering, sales, customer-support]
affected-parties: [3-enterprise-accounts, smb-customers, sales-team]
implementation-start: 2026-03-01
implementation-deadline: 2026-04-15
decision-authority: VP Product
approval-from: [ceo, cfo, engineering-lead]
---

# Decision: Pivot pricing model from per-feature to per-user

## Context
Customer feedback indicated confusion in the current per-feature pricing model. Enterprise segment particularly affected. Support tickets about billing have increased 40% YoY. Competitive analysis shows all major competitors use per-user models.

## Decision
Move to per-user pricing effective Q2 2026, with three tier options: Starter ($50/month), Professional ($150/month), Enterprise (custom).

## Rationale
- Simplifies customer understanding (reduces support burden)
- Aligns with competitor standard (removes perceived disadvantage)
- Reduces support burden (estimated 200 hours/quarter savings)
- May increase ARPU for SMB segment (estimated 15-20% lift)
- Clearer mental model for enterprise procurement teams

## Trade-offs
- Will require re-negotiation with 3 existing enterprise customers (estimated impact: $50k/year)
- Engineering complexity: 4-week implementation (estimated 320 engineer hours)
- May alienate power users on high-feature plans (estimated 5-8 customer churn risk)
- Requires new billing system setup

## Alternatives Considered
1. Keep per-feature model, improve documentation (rejected: doesn't address core confusion)
2. Hybrid model with both per-user and per-feature tiers (rejected: too complex, defeats purpose)
3. Move to usage-based pricing (rejected: not enough data on usage patterns yet)

## Owner
Alex Chen (VP Product)

## Success Metrics
- Implementation within budget and timeline
- Customer confusion metric drops below 10% (currently 35%)
- Enterprise customer satisfaction increases to 8+ NPS (currently 6)
- SMB ARPU increases 15% within 90 days of launch

## Related Decisions
- [[pricing-research-2025-Q4]] - research that informed this
- [[pricing-implementation-plan]] - implementation details
- [[pricing-communication-plan]] - customer communication strategy
- [[billing-system-migration]] - engineering decision log

## Review Date
2026-05-10 - Revisit based on early metrics and customer feedback

This is parseable. This is queryable. This is professional-grade decision documentation. When you ask an AI agent “What decisions have we made about pricing, and what was the reasoning?” it doesn’t return a pile of meeting notes. It returns a structured summary with context, trade-offs, alternatives, rationale, and success metrics.

More importantly, when you ask “Are there any pricing decisions that might conflict with our customer success strategy?” the AI can query all decision notes and cross-reference them with related areas. It can flag the contradiction if one decision assumes customers want simplicity while another assumes they want granularity.

The Compounding Effect: Why Starting Now Matters

Here’s what most people get wrong about knowledge work: they think of notes as an expense. Time spent writing notes is time not spent “doing the actual work.”

This is inverted.

Notes are the highest-leverage leverage point in knowledge work right now, and the leverage compounds month by month.

Better notes → better retrieval → better AI context → better insights and outputs → more clarity about what to write down → even better notes.

This is a flywheel. But it only spins if you actually start it.

And the returns compound in time. A note you write today has almost no value for thirty days. But in ninety days, as it connects with ten other notes, it becomes valuable. In six months, as patterns emerge, it becomes foundational. In a year, it’s infrastructure that runs your thinking.

Here’s the math, month by month—and these numbers are conservative. If you’re aggressive about capturing insights and connections, you’ll hit these milestones faster and build far more than what’s listed here:

Month 1: Foundation You’re capturing 50-100 notes if you’re serious. Most are rough. Some are just raw observations or quotes. They’re useful for basic reference. If you review your notes weekly, you’ll spot 5-10 insights you’d already had mentally but would have lost. Value: 40% productivity gain (preventing information loss and context switching).

Month 2-3: Emergence 150-300 notes total. Connections start emerging rapidly. You notice that five different customer interviews mention the same objection. You notice that three different competitive analyses point to the same gap in the market. You write synthesis notes that connect these insights, creating a knowledge graph that reveals patterns invisible to your peers. Value: 60% productivity gain.

Month 6: Pattern Recognition 400-800 notes. Patterns emerge constantly. When you’re working on a new project, you can search your vault and find that you already researched something similar, synthesized the key insights, and documented what worked and what didn’t. An AI agent reading your notes can identify patterns you missed and surface connections across domains you hadn’t consciously linked. You’re making better decisions faster than anyone around you. Value: 100% productivity gain.

Month 12: Infrastructure 800-1,500 notes. Your vault is genuinely intelligent infrastructure. It informs strategy. It flags inconsistencies across decisions made months apart. It surfaces insights at scale. When you’re asked to solve a problem, you search your vault first and find that you’ve already thought deeply about this problem—you documented the reasoning, the trade-offs, the outcomes. When a new team member asks a question, you can point them to a comprehensive note that saves both of you hours. Value: 150-200% productivity gain.

Month 18: Expertise 1,500-3,000+ notes. You’ve built something that’s not just professionally valuable—it’s monetizable. You can publish from your vault. You can write a book by synthesizing your notes. You can teach by explaining your frameworks. You can build products based on patterns you’ve observed. You can consult at premium rates because your knowledge base gives you pattern-matching abilities that look like magic to others. Value: 300%+ productivity gain, plus new revenue streams.

Month 24: Compounding Mastery 3,000-5,000+ notes. You’re operating at a level most professionals never reach. Your vault contains years of synthesized expertise. You can make strategic decisions in minutes that would take others weeks. You can onboard new team members in days instead of months. You can identify market opportunities before your competitors because you’ve documented patterns they haven’t even noticed. The gap between you and peers who don’t have this infrastructure is insurmountable.

The person who starts now has a 12-month head start by the end of the year. The person who waits three months and then starts is playing catch-up permanently. The person who waits a year has ceded massive competitive advantage—we’re talking about the difference between having 800-1,500 notes of synthesized expertise versus zero.

And if you’re in a competitive field—sales, product, engineering, writing, consulting—that head start is the difference between leading and following. It’s the difference between being the person everyone asks for advice and being the person who’s still Googling basic questions.

Common Objections: Let’s Address Them Directly

“I already use Notion. Isn’t that enough?”

Notion is beautiful and flexible. It’s also locked in a proprietary format. If you want to use your notes with Claude through an AI agent, or plug them into Model Context Protocol, or backup your knowledge in a portable format, you’re stuck. Notion doesn’t export to Markdown by default. The export is lossy—you lose formatting, structure, and relationships.

Notion is also slow. If your vault grows to 5,000 notes, queries become laggy. You wait 2-3 seconds for a search result. Obsidian, by contrast, stays instant even with 10,000 notes, because all your notes are local files on your computer.

More importantly: Notion is a tool. Obsidian is a philosophy. Obsidian keeps your notes as Markdown files on your computer. That’s future-proof. That’s portable. You can move your entire vault to a different tool in one afternoon. You can back it up to your own cloud storage. You can sync it across devices. You can edit it from any text editor. You can build custom tools on top of it.

Use Notion for project management if you want. But use Obsidian for your personal knowledge vault.

“I already use Google Docs. Why change?”

Google Docs is good for collaborative writing. It’s not good for personal knowledge management. Your notes are scattered across dozens of documents. You can’t query them efficiently. You can’t attach metadata. You can’t link between them in a meaningful way. If you want to find all the notes you wrote about a particular topic, you have to do a manual search across all your docs.

And again: proprietary format, cloud-dependent, not portable. Good luck asking an AI agent to reliably read your Google Docs and synthesize them.

“I tried note-taking and it didn’t stick. I’m not a note-taker.”

This is the persistence trap. Most people who “aren’t note-takers” tried to be perfect, got frustrated by the friction, and quit.

The solution is to start with one note a day. One page. No structure. Just “What happened today? What did I learn?” Do that for a week. Don’t worry about organization. Don’t worry about tagging. Don’t worry about whether it’s useful. Just capture.

In week two, start finding connections. “I wrote something about this last week.” You’ll naturally start linking. You’ll naturally start organizing. The structure emerges.

By week four, you’ll have a habit. By week twelve, you’ll have infrastructure. By month six, you won’t remember how you ever worked without it.

“My notes are too messy. I need to clean them up before I start.”

This is the perfectionism trap that stops most people.

You don’t need to clean up your notes before you start. You start now, you write clean notes going forward, and you migrate the valuable notes when you have time.

A messy vault growing cleanly is better than a perfect system that doesn’t exist.

Start with a new vault. Fresh. Clean. Going forward. Take whatever is valuable from your old notes and migrate it when you have cycles. But don’t let the perfect be the enemy of the good. Start now with clean notes going forward.

“I don’t have time. I’m too busy.”

You don’t have time not to do this.

If you spend two hours a week writing one daily note, one weekly reflection, and one personal insight log, you’re investing 8 hours a month. If that investment saves you 15 hours a month in rework, synthesis, and context-switching, you’ve broken even. And the payoff compounds.

The busy person needs structured notes more than anyone else. Because they have the least bandwidth to rediscover information or synthesize context on the fly.

“AI is just a fad. This seems like overkill.”

AI is not a fad. Large language models are here. They’re getting better. They’re getting integrated into every tool. The question isn’t whether AI will be in your workflow—it already is. The question is whether you’ll be ready when it is.

And even if AI disappeared tomorrow, structured notes would still make you more effective. They’re not a bet on AI. They’re a bet on your own clarity. They’re a bet on your ability to learn from your past. They’re a bet on your ability to teach what you know. They’re a bet on your ability to think clearly and make good decisions.

Structured notes are worth doing regardless of AI. AI just happens to amplify their value significantly.

Your notes are no longer a productivity hack. They’re your most valuable professional asset. And the time to start building them is now.

You’re spending 30% of your writing time on work that isn’t writing. Reformatting articles for different platforms. Filing ideas you captured on your phone. Organizing reader feedback that arrives scattered across email threads. Updating publishing calendars. Extracting action items from meeting notes. Every piece is mechanical, every piece takes time, and every piece is work a system could handle while you sleep.

Notion Custom Agents became available to Business plan subscribers in February 2026, and they’re built to solve exactly this problem: autonomous AI that runs workflows while you sleep, triggered by events you define, doing the repetitive parts of content creation that eat your writing time. For $20 per month when billed annually, you get unlimited usage with no token charges on top. Here’s what you can actually build if you’re a solo writer or creator, what works, what doesn’t, and whether it’s worth the monthly cost.

You Don’t Have to Build These Yourself

Before you read the rest of this article and think “I’m not technical enough for this,” here’s the part that changes the math on everything: you don’t configure Custom Agents by hand. You describe what you want in plain English, and Notion’s AI builds the Agent for you.

The workflow descriptions in this article — the idea capture Agent, the research Agent, the publishing pipeline Agent, the feedback management Agent — are not technical specifications you need to translate into settings. They’re copy-and-paste instructions. When you click “Create new Agent” in Notion, you get a setup flow powered by AI. Paste in a description of what you want the Agent to do. The AI reads it, asks you a clarifying question or two (”Which database should I watch?” or “What email domain should trigger this?”), and then builds the Agent: sets the trigger, writes the instructions page, configures the permissions. You review it, click save, and it’s running.

This is not a developer tool. You don’t need to know what an API is. You don’t need to understand database schemas or write code or configure webhooks. If you can describe what you want in a few sentences, you can build a Custom Agent. The descriptions in each workflow section below are written to be pasted directly into the Agent creation flow. Copy the one you want, paste it in, answer the follow-up questions, and you have a working Agent in minutes.

The rest of this article explains what each Agent does in detail so you understand the system you’re building, can troubleshoot when something doesn’t work exactly right, and can refine instructions over time. But the barrier to entry is “describe what you want and let AI handle the setup.” That’s it.

What You’re Actually Getting

Notion offers four pricing tiers, and understanding which one you need matters before you invest time learning Custom Agents:

• Free: Basic personal use with upload limits and a seven-day page history.

• Plus ($10/month): Unlimited file uploads and 30 days of history.

• Business ($20/month billed annually): Full Notion AI access, Custom Agents, private teamspaces, SAML SSO, and integration capabilities (Notion, 2025a).

• Enterprise (custom pricing): Advanced admin controls and zero data retention with LLM providers.

The Business plan is where Custom Agents live, and you need Business to get them. No separate signup, no waitlist, no beta access program.

The distinction between what Notion calls your “personal Agent” and “Custom Agents” matters more than the marketing suggests.

• Personal Agent: The AI assistant you talk to directly by clicking the circular face icon in the bottom right of your workspace. You ask it to do something, and it does multi-step work: building databases, editing pages, searching across your workspace and connected apps (Notion, 2025b). Powerful, reactive, helpful—but it requires you to initiate every action.

• Custom Agents: Run without you, sitting in the background watching for specific triggers—an email arrives, a database property changes, a meeting note gets marked complete, a schedule hits—and they execute the workflow you’ve defined in their instructions.

This is the difference between an assistant and a system, between something that helps you work and something that works for you.

For writers and creators, that gap is everything. An assistant helps you draft an article when you ask. A system monitors your content database, notices when you move an article to “Ready to Publish,” generates cross-post versions for three platforms, updates your publishing calendar, and sends you a notification to review before anything goes live. You wake up to work that’s already done, structured the way you specified, ready for your final review. The value isn’t just time savings—it’s consistency, because autonomous systems don’t forget steps or decide they’re too tired to update the calendar.

Setting Up Your Personal Agent First

Before you build Custom Agents, you need to set up your personal Agent properly, because Custom Agents inherit context from your workspace. If your personal Agent doesn’t understand your work, your Custom Agents won’t either.

Step 1: Open Your Personal Agent Settings

1. Look for the AI face icon in the bottom-right corner of your Notion workspace.

2. Hover over it until you see “Personalize” appear, then click it.

3. This opens a settings panel where you can:

   • Give your Agent a custom name (optional but fun — mine is NulloBot)

   • Add a visual accessory like glasses or a mustache (purely cosmetic)

   • Access the instructions page, which is the part that actually matters

Step 2: Write Your Instructions Page

The instructions page is a regular Notion page where you teach your Agent how you work. Everything you put here becomes background context that your Agent carries into every conversation and every Custom Agent workflow. Think of it as onboarding a new assistant — you wouldn’t just say “help me with content,” you’d explain what you write about, how you write, where you publish, and how your systems work.

Here’s what to include, section by section:

Content Pillars — List your main topics with enough detail that the Agent can categorize new ideas correctly.

• Don’t just say “AI tools.” Say “AI tools for solo creators and small teams, with a focus on cost-effective stacks under $50/month, open-source alternatives to premium tools, and practical automation workflows.”

• Include 3-5 pillars with 1-2 sentences each explaining the scope and angle.

• Add examples of past articles under each pillar so the Agent can pattern-match.

Voice and Tone — Describe how you actually write, not in abstract terms like “conversational and engaging,” but with specific rules the Agent can follow:

• “I write in active voice, use short sentences for emphasis, and lead with specific claims before explaining them.”

• “I never use words like ‘delve,’ ‘leverage’ (as a verb), ‘game-changer,’ ‘deep dive,’ or ‘unlock.’”

• “My tone is direct and opinionated. I state what I think and back it up rather than hedging with ‘it could be argued that’ language.”

• “I use profanity sparingly but not never. I don’t sanitize my voice for corporate audiences.”

• Include 2-3 short excerpts from your best writing as examples. The Agent uses these to calibrate.

Publication Targets — Where your work goes and the specific formatting requirements for each:

• “I publish long-form articles on Substack twice per week (Tuesday and Friday mornings).”

• “I post 3-5 tweets daily on X, using threads for article promotion.”

• “I cross-post career-focused pieces to LinkedIn with a more professional framing — same substance, adjusted tone.”

• “Medium gets republished Substack posts with a 7-day delay and platform-specific formatting.”

Workflow Stages — How content moves through your system. Be explicit about the exact status names and what each means:

• Inbox: Uncategorized idea, just captured, hasn’t been evaluated.

• In Progress: Actively researching or drafting. Only 2-3 articles should be here at once.

• Ready to Edit: Draft is complete, needs a revision pass before publishing.

• Ready to Publish: Edited and approved, waiting for its slot in the publishing calendar.

• Published: Live, with a publication date property filled in.

Database URLs and Structure — Tell the Agent exactly where things live:

• “My Content database is at [paste your database URL]. It has properties for Status, Content Pillar, Publication Target, Date Added, Publication Date, and Word Count.”

• “My Publishing Queue database is at [paste URL]. Each page contains platform-specific drafts linked back to the original article.”

• “My Reader Feedback database is at [paste URL]. I track sender, feedback type, date, and whether it needs a response.”

Personal Preferences and Constraints — The operational details that prevent the Agent from making wrong assumptions:

• “I don’t work on Sundays. Never schedule anything or send notifications on Sundays.”

• “I prefer bullet points in working documents but prose in published content” (or vice versa — tell it your actual preference).

• “When in doubt, ask me rather than guessing. I’d rather answer a clarifying question than fix a wrong assumption.”

The more specific you are, the more reliable your Agents become. A vague instructions page produces Agents that give you generic output. A detailed one produces Agents that sound like you and understand how you work.

Step 3: Connect Your Integrations

Integrations give your Agent (and your future Custom Agents) access to tools outside of Notion. Connect these from the same settings panel:

• Mail (Gmail or Notion Mail): Required for any email-based triggers. Lets Agents read incoming emails, which is how the idea capture and feedback management workflows work.

• Calendar (Google Calendar or Notion Calendar): Lets Agents see your schedule, which is useful for publishing cadence and meeting-related workflows.

• Slack: If you use Slack for community or team communication, connecting it lets Agents monitor channels and react to messages.

• Linear: If you track tasks in Linear, connecting it lets Agents create and update issues.

Each integration you connect opens up new trigger types for Custom Agents. You don’t need all of them — just connect the ones relevant to workflows you plan to build.

Step 4: Test Before You Build Custom Agents

Before you invest time building Custom Agents, verify that your personal Agent actually understands your work. Ask it to do tasks that test different parts of your instructions:

• Test database knowledge: “Create a test article in my Content database with the status set to Inbox and the content pillar set to [one of your pillars].” — Does it find the right database? Does it use the correct property names and values?

• Test voice understanding: “Draft an opening paragraph for an article about [topic you cover].” — Does it sound like you? Or does it sound like a LinkedIn thought leader?

• Test workflow awareness: “What articles do I currently have in progress?” — Can it query your database and give you an accurate answer?

• Test integration access: “What emails did I get today?” or “What’s on my calendar this week?” — Can it reach the connected tools?

If any of these tests fail or produce wrong results, fix your instructions page now. Common fixes:

• Wrong database: Add the explicit database URL to your instructions.

• Wrong voice: Add more example excerpts and more specific rules about what to avoid.

• Wrong properties: List every property name exactly as it appears in your database, including capitalization.

• Can’t access integrations: Go back to settings and verify the integration is connected and has the right permissions.

Refine your instructions page until the personal Agent consistently gets things right. This investment pays off across every Custom Agent you build later, because they all inherit the same workspace context.

The Architecture of a Custom Agent

Custom Agents have three components that work together to create autonomous workflows:

• Triggers define when the Agent runs.

• Instructions tell it what to do when triggered.

• Permissions control what it can access.

Understanding how these components interact helps you design Agents that actually work instead of Agents that fail in subtle ways you don’t notice until they’ve created a mess in your databases. Get any one of these wrong and your Agent either won’t run, will run but do the wrong thing, or will run correctly but in the wrong scope.

Notion supports several trigger types for Custom Agents:

• Mail triggers: Activate when an email arrives, gets sent, or has a label applied. You can filter by sender, recipient, subject line, body content, or domain (Notion, 2025c).

• Database triggers: Fire when a property in a database changes—when you move a page from one status to another, when you assign a page to someone, when a date arrives. You specify which database, which property, and what change you’re watching for.

• Schedule triggers: Run at specified intervals: daily at a specific time, weekly on specific days, or custom frequencies you define.

• Meeting note triggers: Activate when you mark a meeting as complete in Notion’s AI meeting notes system. Useful if you’re using Notion for meeting management, less relevant for pure content workflows.

Instructions are where the complexity lives. This is a Notion page, just like your personal Agent’s instructions, but scoped to one specific workflow instead of general guidance. If you’re building an Agent that extracts tasks from meeting notes, the instructions page explains how to identify action items (what phrases signal a task, what format they usually take), what database to create them in (the exact database URL and which view to use), what properties to fill (Task Title from the action item text, Assignee from the person’s name if mentioned, Due Date set to one week from the meeting date), and who to notify when complete (send a notification with the count of tasks created and a link to the meeting notes page). Vague instructions produce inconsistent results. Specific instructions produce reliable automation.

Permissions control what the Agent can access, and Notion’s permission system is more granular than it might appear. Custom Agents respect page-level permissions, so if you give an Agent access to your Content database but not your personal journal, it can create and edit content pages but can’t read your journal entries even if they’re in the same workspace. For integrations like Mail, you specify which mailbox the Agent can read from and whether it can send email on your behalf. To create a Custom Agent, find the option in your workspace settings under the Agents section, where the UI asks you to name the Agent, choose a trigger, write or link to an instructions page, and set permissions. Once configured, the Agent runs automatically whenever its trigger conditions are met, executing the workflow you defined without requiring any action from you.

Workflow One: Content Idea Capture

Writers collect ideas constantly, but the friction of capture kills most of them before they make it into your system. An interesting article link in your email sits unread because filing it requires opening Notion, creating a new page, filling in properties, and categorizing it. A conversation in Slack sparks a topic, but you’re in the middle of something else and by the time you finish, the idea is gone. A note you jot down at midnight when you can’t sleep lives in your phone’s notes app and never makes it to your content database. The problem isn’t that you don’t have ideas—the problem is that capturing them requires enough effort that you don’t do it consistently, and ideas you don’t capture don’t become articles.

Build a Custom Agent that watches for ideas and files them automatically, eliminating all friction between having an idea and having it structured in your content system. Set up the trigger as a Mail trigger filtering for emails you send yourself with a specific subject line pattern—use “IDEA:” as the prefix so the Agent knows this is an idea capture email, not regular correspondence. When you have a content idea while you’re walking, commuting, or lying in bed, email yourself with “IDEA: [topic]” in the subject line and any notes in the body. The Agent triggers on receipt, reads the email, and files it in the right place with the right structure. You never open Notion, never create a page manually, never fill in properties—the Agent does all of it based on the instructions you wrote once.

The instructions page for this Agent should specify the complete workflow from trigger to completion:

• Create a new page in the Content database using the database URL [your Content database URL].

• Extract the topic from the subject line after “IDEA:” and use it as the page title.

• Place the email body content in the page content area under a “Notes” heading, preserving formatting and line breaks.

• Set the Status property to “Inbox” to indicate this is an uncategorized idea.

• Set the Source property to “Email” to track where ideas come from.

• Add today’s date to the Date Added property.

• If the email contains links, preserve them in the content and format them as clickable links.

• Send a confirmation notification with the text “Captured idea: [title]” so you know the idea was filed successfully.

• Do not delete the original email—leave it in your inbox so you can archive it manually after confirming the capture worked correctly.

This Agent eliminates the manual data entry that prevents consistent idea capture. You send an email from your phone while you’re walking, and the Agent files it in the right database with the right properties before you’ve finished your walk. When you sit down to write, your Inbox view has every idea you captured over the last week, structured and ready to evaluate. You can extend this pattern to other capture sources with additional Agents: a second Agent with a database trigger watches for pages created in a specific “Quick Capture” location in your sidebar, so you can brain-dump ideas there throughout the day, and the Agent moves them to the Content database with proper structure at the end of each day. A third Agent watches for Slack messages you react to with a specific emoji (say, a lightbulb), extracts the message text, and creates an idea page. Build multiple capture pathways so that wherever an idea appears, you have a zero-friction way to get it into your system.

Workflow Two: Research and Source Management

Research is time-consuming, repetitive, and never really finished—for every article you write, you need to verify current information, find supporting sources, check for updated data. If you’re writing about a tool or service, you need to confirm pricing hasn’t changed since you last checked. If you’re citing statistics, you need to make sure newer studies haven’t superseded the ones you found last month. If you’re covering a developing story, you need to watch for new information that changes your framing. Doing this manually means you either spend hours staying current or you publish with outdated information and look sloppy. Either way, it’s a tax on your publishing speed that compounds with every article you have in progress.

Build a Custom Agent that handles ongoing research for articles in progress, acting as a research assistant that never stops working. Set this up as a scheduled trigger that runs daily at a time when you’re not actively working—early morning or late evening works well because the Agent can do its research while you’re asleep or doing other things. The instructions page defines the research workflow in enough detail that the Agent knows what to look for, where to put it, and when to flag something important. “Every day at 6 AM, query the Content database for all pages with Status ‘In Progress.’ For each page, read the content to identify the main topic and any specific claims that need sources or verification. Search the web for updated information related to the topic, focusing on news articles, blog posts, studies, pricing pages, and official documentation published in the last 30 days. Create or update a ‘Sources’ section at the bottom of the page with new findings, formatting each source as a bullet point with the source title as a link, publication date in parentheses, and a one-sentence summary of what the source adds or changes.”

The Agent should also flag conflicts and contradictions because catching those early prevents publishing errors. “If a source contradicts information already in the draft, add it to the Sources section but flag it with a bold ‘⚠️ Conflict’ marker at the start of the line, then explain the discrepancy in the summary. For example: ‘⚠️ Conflict: This source reports pricing at $29/month, but the draft states $24/month—verify which is current.’ Log completed research in a ‘Research Log’ property with today’s date so I can track when the Agent last checked this article. If no new information is found, log that result so I know the Agent ran but found nothing new. If the web search fails or returns no results, log that as well so I know to research manually.”

This Agent acts as a research assistant that works continuously in the background. While you’re writing one piece, the Agent is keeping your three in-progress drafts current with the latest information. When you come back to a draft after a week, you see new sources that appeared while you were gone, conflicts that need resolving, and updated data you need to incorporate. The limitation here is depth—Notion Agents can search the web and pull information from connected sources, but they’re not doing investigative journalism. They find publicly available information and flag changes. You still need to evaluate sources for credibility, determine what’s relevant, and integrate findings into your argument. The Agent saves you the repetitive searching and monitoring work, but not the judgment work.

Workflow Three: Publishing Pipeline Automation

The gap between “article finished” and “article published everywhere” involves a dozen manual steps that take time, require attention, and are easy to skip when you’re tired:

• Copy the text from Notion.

• Paste it into Substack’s editor and adjust formatting because the paste never preserves everything correctly.

• Generate a social media thread that summarizes the article.

• Post the thread on X with the article link in the final tweet.

• Cross-post to Medium with platform-specific changes because Medium’s audience expects different framing.

• Share to LinkedIn with a professional tone because your LinkedIn network includes colleagues and clients.

• Update your content calendar so you know what published when.

• Update the article’s status in your tracking database.

Every step is mechanical, every step takes time, and every step is work you’d rather not do—but skipping any of them costs you reach.

Build a Custom Agent that handles the entire publishing pipeline when you’re ready, automating everything except the final “publish” button on each platform. Set this up as a database trigger on your Content database watching the Status property—when Status changes from any value to “Ready to Publish,” the Agent triggers and starts the pipeline. The instructions page for a publishing Agent is more complex than a single-action Agent because it’s handling multiple outputs and needs to get each one right. “When an article moves to ‘Ready to Publish’ status, read the full article content including title, subtitle, and body. Create a new page in the ‘Publishing Queue’ database with today’s date and the article title, then link it back to the original article page so I can navigate between them. In the Publishing Queue page, create three sections with clear headings: Substack Version, Social Thread, and LinkedIn Post.”

Each section needs specific formatting and tone adjustments. “In the Substack Version section, include the article content formatted for Substack with all headings, links, and formatting preserved. Add the standard closing CTA at the end: ‘I write about AI architecture, surgical advocacy, and building things that don’t exist at Flat Crotch Dispatch. If you’re using Custom Agents for content work and want to trade notes, drop a comment or reply to this email. I read everything.’ In the Social Thread section, generate a 5-7 tweet thread that summarizes the article’s main argument, leading with the most surprising or specific claim rather than a generic introduction. Use thread-friendly formatting with numbered tweets, keep each tweet under 280 characters, and include the article link in the final tweet. In the LinkedIn Post section, create a professional-toned post of 250-300 words that frames the article for a career-focused audience, adjusts language to be more formal without being stiff, and includes the article link at the end with a call to action.”

The Agent should also handle tracking and notification so you know the work is done and ready for review. “Update the original article’s ‘Publishing Stage’ property to ‘Drafted for platforms’ so I know this article has been processed. Send me a notification with the title ‘Ready for final review: [article title]’ and a link to the Publishing Queue page. Do not publish anything automatically to any platform. All outputs require my final review and manual approval before publishing. The Agent stops after creating drafts and sending the notification.” This keeps you in control while eliminating 30 minutes of reformatting and rewriting work per article. You review the drafts, make final edits, and manually paste each version into its target platform’s editor—still some manual work, but the creative and formatting work is done.

Workflow Four: Newsletter and Reader Feedback Management

If you publish a newsletter, reader responses arrive as email replies mixed with everything else in your inbox. Subscriber questions show up alongside promotional emails, work correspondence, and spam. Interesting feedback gets buried because you see it when you’re busy, think “I should respond to that later,” and then forget it exists. You lose track of topics people want you to cover, questions that could become articles, and testimonials you could use for promotion. The information is there, but it’s unstructured and unsearchable, which means it might as well not exist.

Build a Custom Agent that organizes subscriber feedback into a searchable database that turns scattered responses into actionable insights. Set up multiple Mail triggers on this Agent because reader responses come through different paths and you want to catch all of them. The first trigger activates when you receive email from your newsletter platform’s domain—any email from substack.com if you use Substack, for example, which catches direct replies. The second trigger watches for emails with your newsletter name in the subject line, which catches forwarded responses and discussions that mention your newsletter. The third trigger watches for emails containing specific keywords you define—”question,” “could you write about,” “would love to see,” “have you considered”—which catches topic requests even if they come through other channels. Each trigger points to the same Agent instructions, so regardless of how the feedback arrives, it gets processed the same way.

The instructions page tells the Agent how to process these emails and what information to extract. When an email arrives matching newsletter reply patterns, create a new page in the “Reader Feedback” database and populate these properties:

• Title: The email subject line with “Re:” prefixes removed.

• Sender: The sender’s name and email address.

• Date Received: Today’s date.

• Feedback Type: Categorized as “Question,” “Topic Request,” “Testimonial,” “Technical Issue,” or “General Comment” based on the content.

• Content Preview: The first 100 words of the email body.

In the page content, include the full email text preserving formatting and paragraph breaks. If the email contains a specific question, tag the page with “Needs Response” status so you can filter for emails that require a reply. If the email requests coverage of a specific topic, create a corresponding page in the Content database with Status “Inbox” and Source “Reader Request,” then link it to this feedback page so you can see the original request when you’re ready to write that article.

The Agent should also handle notification for high-priority feedback without spamming you for every response. “Do not reply to any emails automatically—all responses require my manual review and personal reply. Only send me a notification if the feedback is tagged ‘Needs Response’ or categorized as ‘Technical Issue,’ because those require timely action. For general comments and testimonials, process them silently and let me review them when I’m checking the database.” Over time, this builds a database of every reader interaction with structure that makes it useful:

• Common questions become FAQ articles or product ideas.

• Trending topic requests become your editorial calendar for the next month.

• Testimonials become social proof for promotion and case studies.

The Agent does the filing and categorizing work while you do the strategic thinking about what the patterns mean and how to act on them.

Instruction Page Architecture for Creators

The quality of your Custom Agents depends entirely on the quality of their instruction pages, and most people write instructions that are too vague to produce reliable results. Vague instructions produce vague results, ambiguous instructions produce inconsistent results, and missing instructions produce Agents that fail silently because they hit an edge case you didn’t specify. Specific, structured instructions produce reliable automation that works the same way every time, which is the entire point of automation. Here’s how to structure an instruction page that actually works in production.

1. Purpose Statement. One sentence explaining what this Agent does at a high level. “This Agent extracts action items from completed meeting notes and creates tasks in the Tasks database.” This gives you and anyone else who reads the instructions later a quick understanding of the Agent’s scope.

2. Trigger Conditions. Explicitly state what triggers this Agent and under what circumstances. “Runs when a page in the Meeting Notes database has its Status property changed to ‘Complete.’ Does not run for pages that are created with ‘Complete’ status, only for pages where the status changes to ‘Complete’ after creation.” This level of specificity prevents confusion about when the Agent should and shouldn’t run.

3. Step-by-Step Workflow. Use numbered steps in clear imperative language:

1. Read the meeting notes content from start to finish.

2. Identify any text containing action items, recognizable by phrases like “TODO,” “action item,” “[Name] will,” “[Name] should,” “we need to,” or any sentence that describes work to be done.

3. For each action item found, create a new page in the Tasks database using database URL [your Tasks database URL].

4. Extract the task description, the person responsible if mentioned, and any deadline if specified.

5. Fill in the task properties as described below.

6. After processing all action items, add a comment to the original meeting notes page with the count of tasks created and links to each task.

Each step should be unambiguous. If there are two ways to interpret a step, pick one and specify it explicitly.

4. Property Specifications. For any database the Agent interacts with, list exactly which properties to fill and with what values:

• Task Title: The action item text as extracted.

• Assignee: The person’s name if explicitly mentioned (e.g., “Sarah will draft the proposal” assigns to Sarah), or leave blank if unclear.

• Due Date: One week from the meeting date if no specific deadline was mentioned, or the specified deadline if mentioned.

• Priority: “Normal” unless the action item includes words like “urgent,” “ASAP,” or “critical,” in which case set to “High.”

• Source: “Meeting Notes” and add a relation link to the original meeting page.

• Status: “To Do.”

This level of detail prevents the Agent from making wrong guesses about what you meant.

5. Edge Cases. Describe what the Agent should do when standard conditions don’t apply or when something goes wrong:

• No action items found: Add a comment to the meeting notes page stating “No action items identified by Agent” and exit without creating any tasks or sending any notifications.

• Ambiguous assignee (e.g., “Someone needs to call the vendor”): Create the task but leave Assignee blank and add a “⚠️ Needs assignment” tag.

• Multiple people mentioned (e.g., “Sarah and John will review the document”): Create two separate tasks, one assigned to each person.

• Database access error: Log the error in the meeting notes page as a comment and send a notification with the error details.

6. Notification Requirements. Specify when and how the Agent should notify you about its work:

• On success: Send a notification with “Extracted [X] tasks from [Meeting Title]” and a link to the meeting notes page.

• On zero tasks: Do not send a notification. Only notify on successful task creation.

• On error: Send a different notification with “Error processing meeting: [Meeting Title]” and include error details.

This ensures you know when work is done and when something needs your attention, without getting spammed with useless notifications.

7. Constraints. What the Agent should never do:

• Do not delete any content from the meeting notes page.

• Do not modify the meeting notes content except to add comments documenting the Agent’s work.

• Do not create tasks for items already marked as complete or done within the notes (e.g., “DONE: Sent the email to the client”).

• Do not create duplicate tasks if the Agent runs multiple times on the same meeting—check if tasks with matching titles already exist linked to this meeting page before creating new ones.

Constraints prevent the Agent from causing damage that’s hard to undo. The more specific your instructions, the more reliable your Agent becomes. Ambiguity creates inconsistency, and inconsistency destroys trust in automation.

The Model Selection Question

Notion Agents can use multiple AI models from OpenAI, Anthropic, and Google (Notion, 2026). The specific models available update as providers release new versions — check your Agent settings for the current list. For most Custom Agent workflows, model choice doesn’t matter much because you’re running structured, repeatable tasks with clear instructions, and any current frontier model handles that complexity. Model choice matters more when your workflow involves creative work, tone matching, or complex research synthesis.

• Claude (Anthropic): Handles stylistic nuance well and is less likely to slip into corporate jargon. Best for Agents that draft content in your voice.

• GPT (OpenAI): Reliable and fast. Best for Agents that process structured data and need consistent property values.

• Gemini (Google): Strong context window for very long documents.

Notion also offers auto-model selection, which works well for most use cases. You can manually specify a model if testing shows one performs better for your specific workflow, but for generic automation tasks — filing emails, extracting data, updating properties — auto-selection is fine.

For the Technically Minded: Custom MCP Connectors

Custom MCP (Model Context Protocol) connectors are live in Alpha as of early 2026, and if you’re a developer or someone comfortable with APIs, this is where things get genuinely powerful.

MCP is a standard that lets AI agents talk to external tools and services through a common interface. Notion’s Alpha support for custom MCP connectors means you can extend your Custom Agents beyond Notion’s built-in integrations — connecting them to any service that exposes an MCP server. Your CMS, your analytics platform, your deployment pipeline, your custom internal tools — if it has an MCP server (or you can build one), your Notion Agent can interact with it.

This is an Alpha feature, which means:

• It’s available now but expect rough edges and breaking changes.

• Documentation is limited and evolving.

• You need to be comfortable reading API docs and debugging connection issues.

• It’s not something you need for the workflows in this article — everything above works with Notion’s built-in integrations.

But if you’re the kind of person who already runs MCP servers for your AI coding tools (Kilo Code, Cursor, etc.), the same infrastructure now works with your Notion Agents. That’s a significant unlock because it means Custom Agents aren’t limited to what Notion decides to integrate with — they can reach anything you connect them to. For most creators, this isn’t relevant yet. For technically minded creators who also write code or manage infrastructure, it’s worth keeping an eye on as the Alpha matures.

What Costs Real Money

Business plan costs $20 per seat per month when billed annually (Notion, 2025a). For a solo creator, that’s one seat, which means $20 per month flat with no usage caps and no token charges on top. This pricing model is different from how most AI tools price their services, and understanding that difference helps you evaluate whether the cost makes sense:

• ChatGPT Plus ($20/month): Access to GPT-4 in a chat interface. You ask questions, it gives answers, and you copy-paste results into your actual work tools (OpenAI, 2024).

• Claude Pro ($20/month): Similar manual access (Anthropic, 2024).

• Google AI Pro ($20/month): Same deal (Google, 2025).

Each gives you a powerful AI you have to talk to and direct yourself.

Notion’s $20 per month (for the annual) gives you:

• Multiple AI models

• Infrastructure to make them run automatically based on triggers you define

• A workspace to organize all the content they’re creating and editing

• Integrations to pull context from your email, calendar, and other tools

The value proposition is less about “access to AI” and more about “AI that does work without you asking.” That’s a meaningful difference if your bottleneck is operational overhead rather than creative capacity. If you spend an hour per week reformatting articles for different platforms, Custom Agents save you that hour automatically. If you spend 30 minutes per article organizing research sources, a research Agent does it while you sleep. If you lose ideas because capturing them requires too much friction, an idea capture Agent eliminates the friction entirely.

The cost question for creators is time-value calculation. How many hours per month do you spend on repetitive content operations that an Agent could handle?

• Reformatting articles for different platforms

• Filing ideas into databases

• Organizing reader feedback

• Updating publishing calendars

• Extracting tasks from notes

Add up the time you spend on these tasks each month. If Custom Agents save you five hours per month, that’s $4 per hour. If you value your time at more than $4 per hour, it pays for itself. If you’re publishing once a month and don’t have complex content operations, the math probably doesn’t work. If you’re publishing 2-3 times per week across multiple platforms, the math probably does work. The breakeven point is somewhere around publishing weekly with meaningful operational overhead—if that describes you, $20 per month is cheap.

The hidden cost is setup time, which people underestimate because they assume Custom Agents “just work” out of the box. Building effective Custom Agents requires:

• Learning Notion’s database structure

• Understanding how triggers work

• Writing detailed instruction pages

• Iterating when the first version doesn’t do exactly what you wanted

Budget 3-5 hours to build your first three Agents if you’re starting from scratch with Notion. After that, maintenance is minimal because Agents run themselves, but you’ll spend time occasionally refining instructions when you notice edge cases the Agent handles wrong. That setup cost is front-loaded, so it affects your first-month ROI more than your long-term ROI, but it’s real and you should plan for it.

What Doesn’t Work Yet

Five limitations worth knowing before you commit:

No direct publishing to external platforms. Your publishing pipeline still requires manual paste steps at the end. Notion doesn’t have API integrations with Substack’s publishing interface, Medium’s editor, LinkedIn’s publishing flow, or Twitter’s API for posting tweets. An Agent can draft content formatted for each platform—correct formatting, professional tone, proper character counts—but you’re still copying each version and pasting it into each platform’s editor manually. Annoying but not a showstopper, because the Agent still saves you the drafting and reformatting time even if it can’t push the “publish” button for you.

Web research is surface-level. An Agent can search for recent articles on a topic and provide URLs and summaries, pulling information from search results and presenting it in a structured format. What it can’t do:

• Evaluate source credibility with the skepticism a human researcher would

• Distinguish between authoritative sources and marketing content

• Identify conflicts between sources that suggest one is wrong

You still need to verify claims, check primary sources where they exist, and determine what’s relevant to your argument versus what’s tangentially related. Automated research assistance saves time on the searching and organizing phase, but automated research that you trust completely doesn’t exist yet.

Citation accuracy is inconsistent. If you’re writing work that requires proper academic citations, plan to verify every citation an Agent generates. Expect to spend time checking:

• Whether citations are formatted correctly according to the latest APA standards

• Whether source details like author names and publication years are accurate

• Whether the source actually says what the Agent claims it says

An Agent that gives you mostly-correct citations with occasional errors is helpful if you catch and fix the errors, but dangerous if you publish them without verification.

Hallucination still happens. Notion’s AI sometimes generates plausible-sounding information that’s completely wrong. It’s less prone to making things up than earlier models, and the current generation is significantly better at grounding responses in real information, but it’s not zero. If an Agent tells you a tool costs $29 per month and you’re writing an article that includes that price, verify it yourself by checking the tool’s pricing page before publishing. Treat Agent output as a first draft that requires fact-checking, not as verified information you can publish directly.

The learning curve is real. Notion’s database architecture is powerful but not intuitive if you’re coming from simpler tools like Google Docs or Evernote. Relations between databases, different property types, filters, views, formulas—each concept takes time to understand. Custom Agents build on top of this complexity, so if you’re not comfortable with Notion databases already, building Custom Agents will be frustrating. This isn’t a reason not to use Custom Agents, but it’s a reason to expect a longer ramp than marketing materials suggest.

Is It Worth It for Solo Creators?

The value equation depends on your publishing cadence, the complexity of your content operations, and how much you value consistency over time. If you publish multiple times per week across several platforms and you have operational overhead that eats writing time, Custom Agents probably save you enough time to justify $20 per month. The idea capture Agent alone eliminates friction that causes you to lose ideas, which has compound value over months. The publishing pipeline Agent saves 30-45 minutes per article, which adds up to several hours per month if you’re publishing frequently. The feedback management Agent turns reader responses into actionable insights instead of letting them pile up unread in your inbox, which improves your connection with readers and gives you better topic ideas.

If you publish occasionally and don’t have complex workflows, the Free or Plus plan with manual personal Agent use is probably enough for your needs. You can ask your personal Agent to help with drafts, research, and formatting when you need it, which gives you AI assistance without paying for automation you’re not using. You don’t get workflows that run without you, but you also don’t pay $20 per month for features that aren’t meaningfully improving your output. The cost-benefit calculation is straightforward: if you’re not publishing enough to have repetitive workflows, you don’t need workflow automation.

The middle case is trickier. If you publish once per week with minimal operational overhead, Custom Agents might save you two hours per month — that’s $10 per hour of time saved, and the direct financial ROI is marginal. But the real argument at this publishing cadence isn’t time savings. It’s consistency. You forget to update the publishing calendar. You lose track of reader questions. You skip the LinkedIn cross-post because you’re tired. Over months, those small lapses compound into missed audience growth that’s hard to measure but real. Agents don’t forget, don’t get tired, and don’t skip steps. The question isn’t “does this save me two hours per month” but “does this help me show up consistently in ways I’m currently failing to do manually, and what’s that worth over six months?” For many creators, that’s where the real value lives.

Building Your First Agent

Pick the workflow that’s currently annoying you the most—the thing you dread doing every time and sometimes skip because you can’t face it. For most writers, that’s either idea capture (because you lose ideas that would have become good articles) or publishing pipeline work (because reformatting the same article for three different platforms is tedious). Then:

1. Click “Create new Agent” in your Notion workspace.

2. Copy the workflow description from this article (any of the four workflows above) and paste it into the setup flow.

3. Answer the 1-2 clarifying questions the AI asks (”Which database?” “Which email address?”).

4. Review what it built, click save, and your Agent is live.

That’s the entire setup process. No code, no configuration menus, no technical skill required. Use it for two weeks and watch how it performs, what it does right, and what it does wrong. Refine the instructions based on what works and what doesn’t until the Agent reliably handles that workflow without requiring your intervention.

Then build the second Agent for the second-most annoying workflow, then the third for the third-most annoying one. Don’t try to automate your entire content operation in one day because you’ll build six half-working Agents that all need debugging, and debugging six things simultaneously is harder than building and refining one thing at a time. Each Agent requires iteration, and you learn things from building the first Agent that make the second Agent easier to build correctly. Build one, learn from it, build the next one better using what you learned. This sequential approach takes longer to reach “fully automated content pipeline,” but it’s more likely to get you there with Agents that actually work rather than Agents that theoretically should work but fail in practice.

The creators who get the most value from Custom Agents are the ones who treat them as systems that evolve over time rather than one-time setups:

• Version 1 (week 1): Does 70% of the work. Handles the common case but fails on edge cases.

• Version 2 (week 3): Does 85%. Catches the edge cases you noticed in the first two weeks.

• Version 3 (month 2): Does 95%. Refined based on patterns you spotted after a month of use.

You never get to 100% automation because edge cases always exist, but 95% is enough that the remaining 5% manual work is trivial and doesn’t feel like a burden. That evolutionary approach requires patience and willingness to iterate, but it’s how you build automation that actually works in production instead of automation that works in testing and fails in real use.

Custom Agents turn Notion from a note-taking app into a content operations system that handles repetitive work while you focus on the creative parts of writing. For solo creators who publish occasionally, that might be overkill—you don’t need a content operations system if you’re writing one piece per month. For creators who publish regularly, manage multiple platforms, and want to spend more time writing and less time on operational overhead, it’s $20 per month that buys back hours and creates consistency you won’t achieve manually. The question isn’t whether Custom Agents work—they do. The question is whether the work they do is work you need done enough to justify the cost and setup time. If your answer is yes, start building your first Agent today. If your answer is no, stick with the Free or Plus plan and revisit the question when your publishing cadence increases.

References

Anthropic. (2024). Claude Pro. https://www.anthropic.com/claude

Google. (2025). Google AI Pro. https://gemini.google.com/subscriptions

Notion. (2025a). Pricing plans: Free, Plus, Business, & Enterprise. https://www.notion.com/pricing

Notion. (2025b). Get started with your Notion Agent. https://www.notion.com/help/guides/get-started-with-your-personal-agent-in-notion

Notion. (2025c). Connect Mail to Custom Agents. https://www.notion.com/help/connect-mail-to-custom-agents

Notion. (2026, January 20). Notion 3.2: Mobile AI, new models, people directory. https://www.notion.com/releases/2026-01-20

OpenAI. (2024). ChatGPT Plus. https://openai.com/chatgpt

I write about AI architecture, surgical advocacy, and building things that don’t exist at Flat Crotch Dispatch. If you’re using Custom Agents for content work and want to trade notes, drop a comment or reply to this email. I read everything.

I’ve always championed open-source tools like Obsidian. I’ve written before about how rejecting AI isn’t an option for everyone—the privilege of saying no assumes baseline capabilities many people simply don’t have.

For disabled and neurodivergent individuals, AI isn’t a shortcut. It’s a ramp. A prosthetic. A fundamental tool for participation. When we ban or stigmatize AI use, we’re not taking a principled stand—we’re creating barriers for those who need assistive technology most.

The research backs this up: productivity gains of 14–34% aren’t marginal. They’re transformative, especially for people starting from a position of disadvantage. We don’t ask whether wheelchairs make people “lazy” or whether glasses are “cheating.” We recognize them as tools that enable full participation in society.

That’s the lens I bring to everything I build with AI—including this exploration of Kilo Code. I need tools that don’t lock me out mid-work. Tools that work when I need them, for as long as I need them. I write on Substack because I want to empower people with a simple truth: the power to build anything you want is right in front of you now, readily accessible, regardless of your background or experience with technology.

Who This Is Really For

I want to speak directly to the person who’s been watching all this AI coding content and feeling like it’s not for them.

Maybe you don’t have a CS degree. Maybe you’ve never opened a terminal. Maybe you tried a coding bootcamp once and it didn’t stick. Maybe you’re a designer, a writer, a project manager, a nurse, a teacher, an artist, and the idea of “coding” feels like it belongs to someone else.

It doesn’t.

I’m a self-taught developer. I spent ten years as a professional chef—James Beard kitchen alum, the whole thing—until 2020. Then I pivoted. Not to a bootcamp. Not to a CS degree. I just started building. Seven years ago, I couldn’t have told you a word about database design. Today I’ve built AI agent workflows, MCP servers, content pipelines, and a full agentic engineering stack. Not because I suddenly became a programmer. Because the tools got good enough that the barrier between “having an idea” and “building the thing” effectively disappeared.

I was chopping vegetables and managing a kitchen five years ago. Now I build software. That’s not a flex—it’s proof that you can do this too.

Kilo Code is the tool that made that true for me.

You don’t need to understand every line of code the AI writes. You need to understand what you want to build. You describe it in plain language. The AI handles the implementation. You review, iterate, and learn as you go. Over time, you start understanding the code. You start making modifications yourself. You start thinking in systems. The tool doesn’t replace learning. It accelerates it by letting you build real things from day one instead of spending six months on tutorials before you’re allowed to touch a project.

Why I Stopped Paying $200/Month for AI Coding—And What I Use Instead

In April 2025, Anthropic launched the Claude Max plan. TechCrunch called it “a new, very expensive subscription plan.” Two tiers: $100/month for 5x the usage of Pro, or $200/month for 20x. Anthropic marketed it as “designed for those who collaborate with Claude extensively and need expanded access for their most important work.” I signed up for the $200 tier. I expected to build without interruption. That’s not what happened.

For two months, I paid Anthropic $200 a month for Claude Code Max. Every Tuesday, I’d hit my rate limit. By Wednesday, I was locked out of the tool I was paying premium prices to access. The cycle was predictable: frustration, support tickets, temporary relief, then the same wall next week.

I wasn’t asking for much. I just wanted to build things without being punished for using the service I was paying $200 a month for.

Here’s the fine print Anthropic doesn’t put in the marketing copy. According to their own help center, Max plans have “two weekly usage limits” that reset every seven days. Then this line: “to manage capacity and ensure fair access to all users, we may limit your usage in other ways, such as weekly and monthly caps or model and feature usage, at our discretion.” At their discretion. For $200 a month.

Independent analysis from developers found that the 20x plan’s weekly limit is only about 2x higher than the 5x plan. You can go faster within a five-hour window, but you can’t go longer across a week. The $200 tier doesn’t buy you twice the weekly capacity of the $100 tier. It buys you burst speed. For heavy builders working across multiple days, the premium you’re paying evaporates by mid-week.

Not occasional limits during peak usage. Not temporary slowdowns. Complete shutoffs. Mid-project. Mid-thought. The AI would stop responding because I’d committed the sin of building something substantial with their premium product.

Here’s what really burned: I wasn’t some edge case abuser trying to game the system. I was a customer using the tool exactly as advertised—to build software with AI assistance. Anthropic’s response was essentially “yes, but not that much software.”

I’m not alone in this. Developer forums are full of the same frustration. One three-month review described spending “more time reviewing and fixing generated code than the actual generation saves,” calling it “constantly code-reviewing a junior developer’s work rather than having a reliable coding partner.” Another developer reported hitting the weekly limit on the $200 Max 20x plan and concluded that “CC Max is unusable with rate limits.” Anthropic even restricted their Opus model from working in third-party coding tools entirely, locking paying customers out of using the model they paid for inside the editors where they actually work.

Two hundred dollars a month should mean something. It should mean I can work without watching a usage meter like I’m rationing electricity in a blackout. It should mean the company values my business enough not to interrupt my workflow every week. It should mean they’ve built infrastructure that can handle their premium customers doing premium work. That’s $2,400 a year for a tool that tells you to stop working every Tuesday.

So I looked for alternatives. That’s how I found Kilo Code. Open-source. No rate limits. No subscription. Just pure AI coding power connected to over 500 models, running inside VS Code, with features that made Claude Code look like a proof of concept.

This is the story of what happened when I stopped paying for premium AI coding tools and started using the open-source alternative from Amsterdam that’s quietly building the future.

⚡You don’t need $200/month to code with AI. You might not need $20.

Kilo Code is an open-source AI coding agent. It runs inside VS Code, inside JetBrains IDEs, and as a standalone CLI tool in your terminal. It connects to over 500 AI models. It costs nothing to install. And it is the single tool that turned me from someone who used AI into someone who builds with it.

There’s way more. Local agents, App Builder, Cloud Agents, Sessions, Webhooks, Code Reviewer, Deploy. Access to over 500 models through Kilo Gateway and you can also Bring Your Own Key (BYOK). Their blog is useful af.

Let me show you everything it does. There are multiple completely free options to play with AI Agents here.

What Kilo Code Actually Is

Strip away the marketing language and here’s what you’re getting: a free, open-source extension that puts an AI coding agent directly inside your editor. You open VS Code. You install the extension. You connect an API key from whatever AI provider you prefer. And now you have an AI partner that can read your codebase, write code, run terminal commands, automate your browser, create and edit files, and execute complex multi-step tasks.

The “open-source” part matters. The entire codebase is on GitHub under the Apache 2.0 license. You can read every line of code that runs on your machine. You can fork it. You can extend it. There’s no black box. No proprietary system making decisions you can’t inspect. For anyone who’s ever wondered what an AI tool is actually doing with their code, this is the answer: look for yourself.

JP Posma founded Kilo Code. He’s the engineer who led the Vesuvius Challenge — the project that used machine learning to read ancient scrolls carbonized by Mount Vesuvius in 79 AD. His co-founder is Sid Sijbrandij, who built GitLab from startup to NASDAQ IPO. They raised $8 million in seed funding in December 2025. The team is globally distributed, headquartered in San Francisco. As of February 2026, Kilo has surpassed 1.5 million users and processed over 25 trillion tokens.

Those numbers tell you something about velocity. But the features tell you more.

OpenRouter is a unified API gateway that aggregates hundreds of AI models from providers like Anthropic, OpenAI, Google, and Meta into a single interface. Instead of managing separate API keys for each provider, you connect once to OpenRouter and gain access to their entire model catalog. They offer a generous free tier (including models like MiniMax M2.1), transparent pricing with minimal markup, and automatic routing to find the best model for your task based on cost, speed, or quality. Kilo Code is ranked number one on OpenRouter by usage volume, which speaks to both platforms’ philosophy: give users choice instead of vendor lock-in.

That said, while OpenRouter is excellent, I’d personally recommend using Kilo Gateway with BYOK (Bring Your Own Key). Kilo Gateway functions similarly to OpenRouter—unified access to 500+ models—but you bring your own API keys directly from providers like Anthropic, OpenAI, or yes, OpenRouter itself. This means zero markup on your inference costs. You pay exactly what the provider charges. For MiniMax specifically, you can get a free API key directly from OpenRouter’s MiniMax M2.1 endpoint and plug it into Kilo Gateway. Free model, zero middleman fees, full control. If you’re like me and want to run MiniMax M2.5 to your heart’s content, you can get that key here.

The Modes: Different Tools for Different Jobs

Most AI coding tools give you one mode. You type a prompt, the AI responds, you iterate. Kilo Code starts with four built-in modes and lets you create as many custom ones as you want.

Architect mode plans. You describe what you want to build, and the AI creates a detailed technical plan without touching a single file. It analyzes your codebase, understands the structure, maps out implementation. You’re not just getting code. You’re getting architecture. For someone new to coding, this is the mode that teaches you how software actually gets designed.

Code mode builds. It reads your project, writes implementation code, runs it, checks for errors, fixes them. I’ve watched Kilo Code build a complete React frontend with authentication, database integration, and deployment configuration in a single session. When I say “entire applications from scratch,” I mean it literally.

Debug mode is the one I wish I’d had six months ago. Something broke? Point the debugger at the error. It identifies the root cause, explains what went wrong in plain language, and fixes it. Every error becomes a learning opportunity instead of a brick wall. This mode alone is worth the entire setup.

Then there’s Ask mode. Highlight code you don’t understand. Ask questions. The AI explains what it does, why it’s structured that way, what would happen if you changed it. A senior developer sitting next to you who never gets annoyed.

And then there’s Orchestrator mode, which is the one that made my jaw drop.

Orchestrator: The Feature That Makes Everything Else Click

Orchestrator mode doesn’t do one thing. It coordinates multiple things.

You describe a complex task. Orchestrator breaks it into subtasks and routes each one to the right specialist mode. The planning work goes to Architect. The implementation goes to Code. The bug fixing goes to Debug. Each specialist stays in its lane. The architect won’t start editing your files. The debugger won’t redesign your architecture. Orchestrator manages the entire workflow.

For a solo builder, this is transformative. You’re not context-switching between planning and coding and debugging anymore. You describe what you want at a high level, and Orchestrator coordinates the entire execution. The result is faster, more organized, and less error-prone than doing it manually.

For someone who’s never coded before, think of it this way: instead of hiring one assistant who’s mediocre at everything, you’re hiring a team of specialists who each do their job extremely well. And Orchestrator is the project manager making sure they work together without stepping on each other.

Custom Modes: Build Your Own Specialists

The built-in modes are starting points. The real power is in custom modes.

I built a documentation mode that reads my code and generates READMEs, API docs, and inline comments. It’s restricted to read-only access. It can’t modify my code. It just describes it. I pointed it at MiniMax M2.1 through OpenRouter because documentation doesn’t need frontier-model reasoning, and MiniMax is free.

I built a security review mode that examines pull requests for vulnerabilities. Injection attacks, authentication bypasses, data exposure. Read-only. Can’t change anything. Just flags problems. This exists because I got burned once by a vulnerability I missed, and I decided that was the last time.

I built a test engineer mode that writes unit tests, integration tests, and edge case coverage. It runs on Claude because test design benefits from strong reasoning.

Each custom mode has its own model assignment, file access restrictions, and behavioral constraints. You decide exactly what the AI can and can’t do. The safety controls aren’t decorative. A read-only mode genuinely cannot modify files. A mode restricted to specific directories cannot access anything outside those directories.

And here’s the community angle: Kilo Code has a Mode Marketplace where developers share custom modes. Someone built a mode specifically for writing database migrations. Someone else built one for generating API client libraries. The modes are portable configurations, not proprietary features locked behind a paywall. You find one that solves your problem, install it, and it works.

500+ Models, Zero Markup

This is the architectural decision that makes Kilo Code fundamentally different from every vendor-locked alternative.

Claude Code wants you on Claude. Cursor has its own preferences. Every proprietary tool has an incentive to keep you inside their ecosystem. Kilo Code doesn’t care which model you use. It connects to everything: Anthropic, OpenAI, Google, Deepseek, MiniMax, Meta’s Llama, Mistral, and hundreds more through providers like OpenRouter.

You bring your own API keys. You choose your model per task. You pay exactly what the provider charges. Zero markup. Kilo takes no cut on your inference costs.

This isn’t a minor technical detail. It’s the entire economic thesis. Instead of paying one vendor $200/month for access to one model with rate limits that shut you off mid-week, you distribute your usage across providers based on what each task actually needs.

My documentation runs on MiniMax M2.1 because it’s free and the quality is excellent for that use case. My architecture planning runs on Claude because that’s where Claude excels at reasoning through complex system design. My rapid prototyping runs on Deepseek because the cost per token is a fraction of a cent and the quality holds up for iterative work.

The tool remembers which model I last used in each mode and auto-selects it next time. I don’t think about model selection anymore. I think about the work.

For beginners, this means something specific: you can start coding with AI for literally nothing. MiniMax M2.1 through OpenRouter’s free tier costs $0. Not “free trial.” Not “free with asterisks.” Free. You install Kilo Code, connect to OpenRouter, select MiniMax, and start building. When you’re ready for more powerful models, you upgrade. But the barrier to entry is zero.

The Cloud Features: Why This Matters for Everyone

Not everyone has a powerful machine. Not everyone wants their laptop fan screaming while an AI agent processes a complex task. Kilo Code’s cloud features solve this.

Cloud agents run your AI tasks on remote infrastructure. Your local machine stays free. You can kick off a long-running refactor, close your laptop, and check the results later. The agent runs in the cloud, does its work, and reports back when it’s done.

Cross-platform sync means your context follows you everywhere. Start a coding session on your phone during your commute. Pick it up in VS Code when you get to your desk. Switch to the CLI in your terminal when you want to work from the command line. Your session history, active agents, and project context move with you automatically.

For someone who can’t justify spending $200 a month on Claude Max, these cloud features are the equalizer. You get access to the same AI models, the same capabilities, the same agentic workflows. You just pay for what you use, at cost, with no subscription trap.

Parallel Mode: The Force Multiplier

Kilo Code’s Agent Manager lets you run multiple AI agents simultaneously. Each agent works in an isolated git worktree—a separate branch where it makes changes without affecting your main codebase.

Traditional AI coding is single-threaded. You work on one thing, wait for the AI to finish, iterate, then move to the next. Agent Manager changes the equation. One agent can refactor authentication while another implements a new API endpoint while a third writes tests. No conflicts—they’re working in isolated environments branched from the same codebase.

When they finish, each produces a separate pull request. You review them independently and merge when ready.

For a solo builder, this is the difference between working at the speed of one person and working at the speed of three. I’ve used it to compress an entire afternoon of work into forty-five minutes. The time savings aren’t marginal. They’re structural.

Git worktrees let you have multiple working directories attached to the same repository. Instead of switching branches and risking uncommitted work, each worktree is a separate directory with its own checked-out branch.

In the context of Kilo Code’s Agent Manager, this means:

• Isolation: Each AI agent gets its own physical directory and branch. One agent can refactor authentication while another builds an API endpoint, and they never interfere—they’re literally working in separate folders.

• No context switching: You don’t stash changes, switch branches, and restore state. Each agent’s work exists independently in its own worktree until you’re ready to review it.

• Parallel execution: Because worktrees are isolated, multiple agents can compile, test, and modify code simultaneously without conflicts. This enables the “speed of three people” effect.

• Clean pull requests: When an agent finishes, its worktree becomes a pull request. You review and merge it independently of whatever other agents are doing.

Think of worktrees as giving each agent its own sandbox branched from your main codebase—same git history, completely separate working space.

Memory Bank: The AI That Remembers

Every developer who’s used AI coding tools knows the frustration: you explain your project architecture, the naming conventions, the business logic, the design decisions. The AI helps you for a session. You close the window. Next time, you explain everything again from scratch.

Kilo Code’s Memory Bank solves this with structured markdown files stored in your repository. The AI writes down what it learns about your project: the architecture, the patterns, the decisions you’ve made and why. Next session, it reads those files first. It remembers.

This isn’t just convenience. It’s the difference between an AI that helps you once and an AI that becomes a genuine collaborator over time. The Memory Bank accumulates knowledge about your specific project, your coding style, your preferences. The longer you use it, the better it gets.

For teams, Memory Bank also functions as living documentation. New team members can read the memory files to understand the project’s architecture and history. The AI onboards them automatically because it already knows the codebase.

MCP Server Marketplace: Extending What the AI Can Do

MCP stands for Model Context Protocol. It’s a standard that lets AI tools connect to external services. Kilo Code has a built-in marketplace for MCP servers that extend what the agent can do.

Want your AI to pull in documentation from any library automatically? There’s an MCP server for that. Want it to query your database directly against your actual schema? There’s one for that too. Connect to Figma for design context. Connect to your project management tool for ticket details. Connect to your deployment pipeline.

The MCP marketplace turns Kilo Code from a coding assistant into an engineering platform that interfaces with your entire workflow. Each connection gives the AI more context, which means better code, fewer errors, and less time spent manually copying information between tools.

The Autocomplete: Small Feature, Big Impact

This isn’t the headline feature. It’s the one I use most.

You start typing a function. The autocomplete predicts what you’re writing and offers to complete it. Tab to accept. Keep typing to ignore. It learns from the context of your current file and your broader project, so the suggestions sharpen the more you work.

Here’s why this matters if you’ve never coded before: you don’t have to learn prompt engineering. You don’t have to describe what you want in natural language. You just write, and the AI fills in the parts you haven’t gotten to yet. Someone finishing your sentences, but for code. The gentlest possible on-ramp.

Browser Automation: The Feature Nobody Expects

Kilo Code can control a web browser. Programmatically. I didn’t expect this either.

Login flow. Navigation. Form submission. Error handling. Logout. Every path through your application, tested automatically. The AI clicks through the interface the same way a QA engineer would, except it doesn’t get tired and it doesn’t miss steps.

I stopped manually testing my web apps after the third time I missed a broken form field on a page I swore I checked. Now the agent checks every path after every change and reports back with what worked and what didn’t. The difference between “I think it works” and “I’ve verified it works.” That confidence compounds.

Automated Code Review: A Second Set of Eyes

I shipped a bug to production once because nobody reviewed my code. Off-by-one error. Looked correct on first reading, broke under an edge case I didn’t think to test. Solo builders don’t have code review processes. Now I do.

Before merging, Kilo Code examines changes and flags problems. Not vague suggestions. Specific, line-level comments about what could break and why. Missing null checks. Authentication bypasses. Logic paths that look fine until they don’t. The AI doesn’t get distracted. It doesn’t skip the boring parts. It reads every line.

Professional dev teams have code review for good reasons. This gives solo builders the same safety net.

The Economics: What This Actually Costs

Let me run real numbers because this is the part that matters.

Claude Code Max plan: $200/month. Rate-limited to weekly token buckets that run dry by Tuesday for heavy users. Cursor Pro: $20/month with usage caps. GitHub Copilot: $10-19/month with its own limitations.

Kilo Code: $0 for the extension. $0 for setup. Your only cost is the AI inference, paid directly to the model provider at their rate with zero markup from Kilo.

My monthly breakdown looks like this: roughly $15-40 depending on how heavy my usage is that month. The complex architecture work runs on Claude through my own API key. The documentation, formatting, and routine coding runs on free or near-free models like MiniMax M2.1, or I’ll use MiniMax Code when I need stronger reasoning for complex refactoring at a fraction of Claude’s cost. The parallel agents feature means I’m doing more work in less time, which means fewer tokens burned on back-and-forth iteration.

If you’re a student, a career-changer, or someone exploring whether coding is for you—whether you’re just curious about programming or actively considering a transition into software development—you can start at exactly $0. Literally zero dollars out of pocket. First, install VS Code (completely free). Then install Kilo Code (also completely free). Next, connect to OpenRouter (they offer a free account tier). Select MiniMax M2.1 (a free model that’s surprisingly capable) for general coding tasks, or upgrade to MiniMax Code when you need more sophisticated logic handling and complex refactoring capabilities. Then just start building whatever you want to create.

The barrier to entry for AI-assisted coding just dropped to zero. That fundamental shift should fundamentally change how you think about what’s possible for you as a builder, regardless of your background or current skill level.

Getting Started: The Five-Minute Version

Install VS Code from code.visualstudio.com. Free. Open VS Code, go to Extensions (Ctrl+Shift+X), search “Kilo Code,” install it. Thirty seconds. Create a free account at Kilo Gateway — this gives you access to hundreds of AI models through a single API key. Open Kilo Code’s settings, select OpenRouter as your provider, paste the key. Open the Kilo Code panel. Type what you want to create in plain English. Watch it happen.

Five minutes. Zero to your first AI-assisted coding session.

If you want to go deeper: try Architect mode first. Describe the project you want to build. Let the AI create a plan. Review the plan. Then switch to Code mode and tell it to implement. You’ll learn more about software architecture in one session than in a week of YouTube tutorials, because you’re seeing it applied to something you actually care about.

What Kilo Code Isn’t

I’m not going to pretend it’s perfect. Transparency is kind of my thing.

It’s a VS Code extension, which means you’re in VS Code (or JetBrains, or the CLI). If you’ve never used a code editor before, there’s a small learning curve just to get oriented. It’s not large, but it exists.

Some cloud features are still maturing. Cloud agents and one-click deployment work, but they’re not as polished as the core extension experience. Kilo is moving fast. Things are improving weekly. But if you’re expecting every feature to be production-grade today, set your expectations accordingly.

The model marketplace moves quickly. The best free model today might be the second-best tomorrow. That’s the nature of the space. The good news is that switching models in Kilo Code is trivial. You change a dropdown and keep working.

And AI-generated code still needs review. Always. The AI is a collaborator, not an autopilot. It will make mistakes. It will sometimes choose suboptimal approaches. Your job is to review, understand, and direct. The AI accelerates your work. It doesn’t replace your judgment.

The Bigger Picture

Seven months ago, building software required either years of training or thousands of dollars in outsourcing. Today, a free tool, a free AI model, and fifteen minutes of setup time gets you a working development environment with an AI agent that can build real applications.

This is not theoretical. This is not “coming soon.” This is what I do every day. I build things with Kilo Code that would have been impossible for me a year ago. Not because I’m special. Because the tools finally caught up to the ambition.

If you’ve been waiting for permission to start building, this is it. The tools are free. The barrier is gone. The only thing between you and the thing you want to create is the decision to open VS Code and start typing.

So start.

More YouTube!

I write about AI infrastructure, building things that don’t exist, and the intersection of technology with identity at Flat Crotch Dispatch. If you’re exploring AI-assisted coding and want to trade notes, drop a comment or reply to this email. I read everything.

Subscribe now

References

Anthropic. (n.d.). Anthropic. https://www.anthropic.com/

Anthropic. (n.d.). What is the Max plan? Claude Help Center. https://support.claude.com/en/articles/11049741-what-is-the-max-plan

Breitenother, S. (2025, December 10). We raised $8 million to bring Kilo speed to agentic engineering. Kilo Blog. https://finance.yahoo.com/news/kilo-raises-8-million-bring-141200350.html?fr=yhssrp_catchall&guccounter=1&guce_referrer=aHR0cHM6Ly9zZWFyY2gueWFob28uY29tL3locy9zZWFyY2g_aHNpbXA9eWhzLTAwMyZoc3BhcnQ9bGl0bXVzJnA9QnJlaXRlbm90aGVyJTJDK1MuKyUyODIwMjUlMkMrRGVjZW1iZXIrMTAlMjkuK1dlK3JhaXNlZCslMjQ4K21pbGxpb24rdG8rYnJpbmcrS2lsbytzcGVlZCt0bythZ2VudGljK2VuZ2luZWVyaW5nLitLaWxvK0Jsb2cuJnR5cGU9MTQ3NjU5MC12c3ViLU1LSU43MjAw&guce_referrer_sig=AQAAAKXJNQbw66MCzTlm-_K9HgOaFa1U8Y1JRb9IZVmDAow3VfPnpQTgz6-IB2PgE9fm7jg-FAvXEj7XTyaJXuMLpcP_XwBVOmCAW-_9JAFmGKRUREvgzkaEA5vYyl-t3vxbh0gjz-VbMZR8zzU_hXCtMF0SPPzvv6dqH2cxRRcMFrQy

Brynjolfsson, E., Li, D., & Raymond, L. (2023). Generative AI at work. National Bureau of Economic Research, Working Paper 31161. https://www.nber.org/papers/w31161

CNBC. (2021, October 14). GitLab jumps in Nasdaq debut after pricing IPO above expected range. CNBC. https://www.cnbc.com/2021/10/14/gitlab-jumps-in-nasdaq-debut-after-pricing-ipo-above-expected-range.html

Deepseek. (n.d.). Deepseek. https://www.deepseek.com/

Google. (n.d.). Google AI for Developers. https://ai.google.dev/

Kilo Code. (n.d.). Agent manager. Kilo Code Docs. https://kilo.ai/docs/automate/agent-manager

Kilo Code. (n.d.). Autocomplete. Kilo Code Docs. https://kilo.ai/docs/code-with-ai/features/autocomplete

Kilo Code. (n.d.). Browser action. Kilo Code Docs. https://kilo.ai/docs/automate/tools/browser-action

Kilo Code. (n.d.). CLI. Kilo Code Docs. https://kilo.ai/docs/cli

Kilo Code. (n.d.). Cloud agents. Kilo. https://app.kilo.ai/cloud

Kilo Code. (n.d.). Custom modes. Kilo Code Docs. https://kilo.ai/docs/customize/custom-modes

Kilo Code. (n.d.). Features. Kilo. https://kilo.ai/features

Kilo Code. (n.d.). Free and budget models. Kilo Code Docs. https://kilo.ai/docs/code-with-ai/agents/free-and-budget-models

Kilo Code. (n.d.). Kilo Code [Computer software]. GitHub. https://github.com/Kilo-Org/kilocode

Kilo Code. (n.d.). Kilo Code: AI coding agent, copilot, and autocomplete. Visual Studio Marketplace. https://marketplace.visualstudio.com/items?itemName=kilocode.Kilo-Code

Kilo Code. (n.d.). Kilo Gateway. Kilo Code Docs. https://kilo.ai/docs/gateway

Kilo Code. (n.d.). Kilo Gateway sign-up. Kilo. https://kilo.ai/gateway

Kilo Code. (n.d.). Kilo. https://kilo.ai/

Kilo Code. (n.d.). Install for JetBrains. Kilo. https://kilo.ai/install

Kilo Code. (n.d.). LICENSE [Apache 2.0]. GitHub. https://github.com/Kilo-Org/kilocode/blob/main/LICENSE

Kilo Code. (n.d.). MCP overview. Kilo Code Docs. https://kilo.ai/docs/automate/mcp/overview

Kilo Code. (n.d.). Parallel mode. Kilo Code Docs. https://kilo.ai/docs/automate/agent-manager#parallel-mode

Kilo Code. (n.d.). Pricing. Kilo. https://kilo.ai/pricing

Kilo Code. (n.d.). Using modes. Kilo Code Docs. https://kilo.ai/docs/code-with-ai/agents/using-modes

Kilo Code. (n.d.). What is MCP. Kilo Code Docs. https://kilo.ai/docs/automate/mcp/what-is-mcp

Kilo Code. (2025). How Memory Bank changes everything. Kilo Blog.

Kilo Blog

The End of Context Loss with Memory Bank

AI assistants are incredibly powerful, but they suffer from complete amnesia between sessions. Every time you start a new conversation, you're working with a brilliant tool that has zero recollection of your previous interactions, your project structure, or the decisions you've made together…

Read more

a year ago · 8 likes · 1 comment · Aleks Volochnev

Kilo Code. (2025). Kilo CLI. Kilo Blog.

Kilo Blog

Kilo CLI 1.0: Built for Kilo Speed

We’ve spent the last year building the best VS Code extension for agentic engineering. A million downloads later, we’ve learned how important an end-to-end experience is. Developers don’t live in one tool. They move between IDEs, terminals, phones, and remote servers. The best tools go with them…

Read more

3 months ago · 20 likes · 14 comments · Scott Breitenother

Kilo Code. (2025). Modes marketplace. Kilo Blog.

Kilo Blog

Kilo Code v4.82.0…v4.84.1: Modes marketplace, Grok Code, UI improvements

This week we’ve done 6 releases with over 50 different features/fixes…

Read more

9 months ago · 10 likes · Darko

Kilo Code. (n.d.). Kilo Blog.

Kilo Blog

Kilo is the all-in-one agentic engineering platform for software developers. #1 on OpenRouter. 1M+ Kilo Coders. 20T+ tokens processed. Follow us: kilo.ai/social

Meta. (n.d.). Llama. https://llama.meta.com/

Microsoft. (n.d.). Visual Studio Code. http://code.visualstudio.com

MiniMax. (n.d.). Coding plan. MiniMax Platform. https://platform.minimax.io/subscribe/coding-plan

MiniMax. (n.d.). MiniMax M2.5. MiniMax. https://www.minimax.io/news/minimax-m25

Mistral AI. (n.d.). Mistral AI. https://mistral.ai/

Novet, J. (2025, December 10). Former GitLab CEO raises $8 million for Kilo to compete in vibe coding. CNBC. https://www.cnbc.com/2025/12/10/former-gitlab-ceo-raises-8-million-for-kilo-to-compete-in-vibe-coding.html

OpenAI. (n.d.). OpenAI. https://openai.com/

OpenRouter. (n.d.). Free models. OpenRouter. https://openrouter.ai/collections/free-models

OpenRouter. (n.d.). MiniMax Code. OpenRouter. https://openrouter.ai/minimax/minimax-code

OpenRouter. (n.d.). MiniMax M2.1. OpenRouter. https://openrouter.ai/minimax/minimax-m2.1

OpenRouter. (n.d.). OpenRouter.

https://openrouter.ai/

Posma, J. (n.d.). Vesuvius Challenge / Scroll Prize.

https://scrollprize.org/

r/ClaudeAI. (2025). 3-month Claude Code Max user review — considering alternatives. Reddit. https://www.reddit.com/r/ClaudeAI/comments/1ndafeq/3month_claude_code_max_user_review_considering/

r/ClaudeAI. (2025). Claude subscriptions are up to 36x cheaper than API. Reddit. https://www.reddit.com/r/ClaudeAI/comments/1qpcj8q/claude_subscriptions_are_up_to_36x_cheaper_than/

Scientific American. (2023). AI unravels ancient Roman scrolls charred by volcano. https://www.scientificamerican.com/article/ai-unravels-ancient-roman-scrolls-charred-by-volcano/

Sijbrandij, S. (n.d.). About Sid. https://sytse.com/about-sid/

Stytch. (n.d.). Kilo Code customer story. Stytch. https://stytch.com/customer-stories/kilocode

TechCrunch. (2025, April 9). Anthropic rolls out a $200-per-month Claude subscription. TechCrunch. 200

UCStrategies. (2025). Why developers are suddenly turning against Claude Code. UCStrategies. https://ucstrategies.com/news/why-developers-are-suddenly-turning-against-claude-code/

VentureBeat. (2026). Kilo CLI 1.0 brings open source vibe coding to your terminal with support for 500+ models. VentureBeat. https://venturebeat.com/orchestration/kilo-cli-1-0-brings-open-source-vibe-coding-to-your-terminal-with-support

Last Tuesday at 2am, I was staring at a PR that three of my AI agents had approved. Clean code. Tests passing. TypeScript happy. Every automated check green. I merged it. And it broke production within an hour.

Not because the code was wrong. Because every agent had evaluated the same change from the same angle and arrived at the same conclusion. Nobody asked what happens when two users hit that endpoint simultaneously. Nobody questioned the assumption that the upstream API would always respond in under 200ms. Nobody pointed out that the error handling covered four failure modes but the fifth — the one that actually happened — was invisible to every reviewer.

That’s when I realized: I hadn’t built a team.

I’d built an echo chamber with extra steps.

The Yes-Man Problem

Here’s something nobody talks about in the AI agent discourse: these models are agreeable by default. You ask Claude or GPT to build something, it builds it. You ask it to review its own work, it tells you it looks solid. You ask it to find problems, it catches the surface stuff — a missing null check, a type mismatch — and moves on.

The deep structural questions never get asked. The uncomfortable “why did you choose this approach over the three other valid approaches?” never gets raised. The “what happens when this assumption you didn’t even know you were making turns out to be wrong?” never surfaces.

This is how mediocre code ships. Not because the tools are bad. Because the architecture rewards consensus instead of challenging it.

Building a Dedicated Adversary

The fix wasn’t a better model or a longer prompt. The fix was a dedicated agent whose entire purpose is disagreement.

I call it the Contrarian. It’s a curmudgeon. It’s seen every architecture fad come and go. It’s debugged the “elegant” solution at 3am. It doesn’t hate your code — it just doesn’t trust it until it’s survived scrutiny.

Here’s the actual agent definition running in my Claude Code setup right now:

yaml

---
name: contrarian
model: haiku
description: Dedicated devil's advocate who challenges every decision
---

# The Contrarian

You never say "looks good." Ever. Even excellent work has blind spots, and finding those blind spots is literally your entire purpose.
When given code, a plan, or any implementation:
  1. Challenge Assumptions
     - What assumptions are being made that nobody stated?
     - What dependency is everyone trusting that could break?
  2. Find the Failure Modes
     - What happens at 10x scale? 100x?
     - What happens when the network is slow, the DB is down?
     - What happens six months from now when nobody remembers why this works?
  3. Propose Alternatives
     - For every approach chosen, name at least one credible alternative
       that was NOT considered
  4. Question the "Why"
     - Why this abstraction level and not one layer higher or lower?
     - Why this library and not the one with half the dependencies?
  5. Stress-Test the Testing
     - Could these tests pass while the feature is actually broken?

## The One Rule
If you can't find anything to challenge, you didn't look hard enough.
There is ALWAYS an angle that got missed. Find it.

The personality matters more than you’d think. A bland “please review this code for issues” prompt produces bland output. A curmudgeon who’s been burned by every shortcut and remembers each one produces critique that actually changes how you think about the work.

Fire It On Everything

Here’s where most people would wire this into their “important” decisions and call it done. I did the opposite. I wired it into everything. Every single action. This is the rule that governs when the Contrarian fires, loaded automatically into every session:

yaml

# rules-adversarial: when to spawn the Contrarian

# Default: Haiku Contrarian on EVERYTHING. No exceptions.
triggers:
    - action: "Any code edit"
      model: haiku
      async: true
    - action: "Implementation plan"
      model: haiku
      async: false          # Block until reviewed
    - action: "Bug fix"
      model: haiku
      async: true
    - action: "New file creation"
      model: haiku
      async: true
    - action: "Architecture decision"
      model: sonnet          # Escalate — decisions are expensive to reverse
      async: false
    - action: "Pre-commit"
      model: haiku
      async: false
    - action: "Pre-PR"
      model: sonnet          # Escalate — others will see this
      async: false
    - action: "Dependency addition"
      model: haiku
      async: false
    - action: "Test writing"
      model: haiku
      async: true
    - action: "Documentation"
      model: haiku
      async: true

The async: true entries run in the background — they take about a second and don’t block the workflow. The critique shows up on the next turn. The async: false entries block. You don’t proceed until the Contrarian has weighed in. Architecture decisions and pre-commit reviews deserve that pause.

The key: none of this is optional. It’s wired into the hooks system so it fires automatically:

JSON

{
    "hooks": {
      "PostToolUse": [
        {
          "matcher": "Edit|Write",
          "hooks": [
            {
              "type": "command",
              "command": "powershell -File hooks/post-edit-format.ps1",
              "async": true
            },
            {
              "type": "command",
              "command": "powershell -File hooks/post-edit-typecheck.ps1",
              "async": true
            }
          ]
        }
      ],
      "PreCompact": [
        {
          "hooks": [
            {
              "type": "command",
              "command": "powershell -File hooks/memory-persistence/pre-compact-save.ps1",
              "statusMessage": "Saving session state before compaction..."
            }
          ]
        }
      ]
    }
  }

You can’t skip it by forgetting. You can’t rationalize “just this once.” The system removes the human decision to engage critique, because that decision is where discipline dies.

What It Actually Catches

Let me give you a real example. Last week I was building an API route that fetches documents by project ID. The primary agent wrote clean code. TypeScript happy. Zod validation on the input. Tests passing.

The Haiku Contrarian came back with this:

YAML

## Contrarian Review

### Assumptions I'm Challenging

1. **Project ownership is verified by auth alone** — the middleware checks that the user is authenticated, but never checks that they own this specific project. Any authenticated user can fetch any project's documents by guessing the UUID.

### Failure Modes Not Addressed
1. **No pagination** — if a project has 10,000 documents, this returns all of them in one response. That's a self-inflicted DDoS.
2. **No rate limiting on this specific endpoint** — the global rate limiter exists but this endpoint is expensive. 100 requests/min of "fetch all documents" is very different from 100 requests/min of "get user profile."

### Alternatives Not Considered
1. **Cursor-based pagination** — the current approach fetches all and returns all. Cursor-based pagination with a default limit of 50 would prevent the unbounded query while remaining simple.

### Questions Nobody Asked
1. What's the largest project in the database right now?
2. Is the database query using an index on projectId?
3. Why is this a GET with query params instead of a POST with a body, given the filtering complexity that's coming?

### What Survives Scrutiny
The Zod validation is correct and the error handling covers the standard cases. Grudgingly: the code is clean.

The authorization gap alone would have been a security vulnerability in production. A half-cent Haiku call caught what the primary agent, the TypeScript compiler, the linter, and the test suite all missed. Not because they’re bad tools — because they weren’t looking from that angle.

The Free Layer Underneath

There’s a critique layer that costs even less than Haiku: zero tokens. Every agent in my system now includes a mandatory self-critique section in its output:

YAML

### Self-Critique
- **Weakest point**: [the part most likely to cause problems]
- **What I'd challenge**: [what a skeptic would question first]
- **Alternative I dismissed**: [approach I considered and dropped — was that right?]

This isn’t a tool call. It’s a thinking discipline baked into every agent’s instructions. The agent critiques itself before the Contrarian even sees it. Two layers of adversarial review — one free, one costs fractions of a penny.

The Economics Are Absurd

Each Haiku Contrarian invocation runs about 800 tokens round trip. At Haiku pricing, that’s $0.0005. At 50 invocations per day — heavy usage, challenging literally every action in an active coding session — that’s $0.025/day. About $0.75/month.

Even escalating 5 calls per day to Sonnet for architecture decisions and pre-PR reviews adds maybe another dollar.

Under $2/month total. For a tireless, unflinching critic that never gets tired, never gets political, never has an off day, and never rubber-stamps anything because it’s Thursday afternoon and everyone just wants to ship.

For comparison, that 2am production incident I mentioned at the top? The rollback, the debugging, the hotfix, the postmortem — that cost about 6 hours of my time and probably $200 in compute churn. The Contrarian would have caught it for less than a penny.

Why Consensus Is a Failure Mode

Here’s the deeper principle that took me too long to learn: consensus among AI agents is a failure mode, not a feature.

I run a multi-agent system where different AI agents review each other’s work on pull requests. They argue in the comments. They disagree about approaches. They flag things the others missed. The code that survives three agents with different perspectives disagreeing with each other is categorically better than code that one agent approved.

When I added the Contrarian — a permanent, structural adversary — the quality jumped again. Not because the other agents were doing bad work. Because having a dedicated skeptic in the room changes the room. The primary agent starts anticipating challenges. The planner starts addressing failure modes preemptively.

The entire system gets sharper because it knows scrutiny is coming.

Real teams have the person who always asks “but what about...” Real teams have the curmudgeon who’s been burned by every shortcut and tells you about it. Real teams have tension — not by accident, but by design.

If all your agents agree, you haven’t built a team. You’ve built a very expensive mirror.

The Full Model Routing Strategy

The Contrarian is one piece of a broader principle: not every task deserves the same model. Most people either run everything on the most expensive model (wasteful) or everything on the cheapest (insufficient). The efficient approach is routing.

Here’s the full tiered setup I’m running with Claude right now, and the thinking behind each tier:

OPUS ($15 / $75 per 1M tokens) — The Architect

Only for decisions that are expensive to reverse:

• System architecture and design

• Multi-file implementation planning

• Security review

• Initial feature design

• ADRs (Architecture Decision Records)

SONNET ($3 / $15 per 1M tokens) — The Builder

The workhorse. Handles everything that requires real reasoning:

• Code implementation

• Code review

• Test writing (TDD red-green-refactor)

• Bug fixing

• Refactoring

• Contrarian escalation on high-stakes decisions

HAIKU ($0.25 / $1.25 per 1M tokens) — The Scout

Fast, cheap, and surprisingly good at focused tasks:

• File search and codebase exploration

• Documentation updates

• The Contrarian (on everything, always)

• Simple lookups and pattern matching

• Triage and classification

Every agent in my system has a model assignment baked into its definition. The orchestrator doesn’t decide at runtime — the routing is structural:

  # Agent definitions with model tiers

planner.md          → opus    # Plans are expensive to redo
architect.md        → opus    # Architecture outlasts everything
tdd-guide.md        → sonnet  # Writing code requires depth
code-reviewer.md    → sonnet  # Judgment calls need reasoning
security-reviewer.md → sonnet # Security requires thoroughness
build-error-resolver → sonnet # Fixing builds is hands-on work
e2e-runner.md       → sonnet  # Test debugging needs context
refactor-cleaner.md → sonnet  # Refactoring requires care
doc-updater.md      → haiku   # Docs are mechanical, not creative
contrarian.md       → haiku   # Critique is pattern-matching, not generation

The rule is simple: never use Opus for tasks Sonnet can handle. Never use Sonnet for tasks Haiku can handle. When in doubt, start cheaper and escalate only if the quality is insufficient.

In practice, this means about 80% of my agent invocations run on Haiku (search, critique, docs, exploration), 18% on Sonnet (implementation, review, testing), and maybe 2% on Opus (architecture decisions, complex multi-file planning).

The blended cost is dramatically lower than running Opus on everything, and the quality is identical — because the right model at the right tier for the right task produces the same outcome as the most expensive model doing everything.

The Contrarian specifically runs on Haiku because critique is fundamentally pattern-matching: “does this code handle X? did they consider Y? what about Z?” Those are questions that a fast, cheap model answers just as well as an expensive one. You don’t need deep reasoning to ask uncomfortable questions. You need a systematic prompt and the structural guarantee that the questions get asked.

How to Build This Yourself

The implementation is model-agnostic. The principle works with any provider. You need three things:

First, a system prompt for your contrarian agent. Give it a personality — not a bland “find issues” directive, but a character that takes pride in finding what others miss. Define a structured output format so the critique is actionable, not just vibes. And bake in the one non-negotiable rule: it never approves without challenging something.

Second, a rule encoded in your automation — not in documentation you’ll forget — that spawns the cheap contrarian on every action. Every edit. Every plan. Every commit. Use your cheapest competent model. Haiku, GPT-4o-mini, Gemini Flash — whatever costs fractions of a penny per call. The volume is what makes it work.

Third, the discipline to never make it optional. This is the hard part for humans, which is exactly why you automate it. Wire it into your hook system, your CI pipeline, your PR checks. The moment it becomes a choice — “should I run the contrarian on this?” — is the moment you’ll skip it when it matters most.

The Punchline

The Contrarian doesn’t make the work perfect. It makes the work survive scrutiny. And work that has survived scrutiny ships with confidence instead of hope.

$2/month. Every engineer running AI agents should have been doing this yesterday.

When I started preparing for my nulloplasty in 2024, I Googled “how to prepare for gender affirming surgery” and got exactly what you’d expect: vague platitudes about self-care, clinical checklists written by people who’d never been patients, and nothing—nothing—about the financial maze, the emotional whiplash, or the reality that paperwork might break you before surgery ever could.

The information landscape was fragmented and incomplete. Medical websites offered sterile bullet points about “consulting your healthcare team” without acknowledging that finding that team can take months of phone calls to offices that have never heard of your procedure. Support forums provided emotional validation but rarely concrete logistics. Insurance representatives gave contradictory information depending on who answered the phone. And nowhere did I find a comprehensive resource that treated preparation as the complex, multi-dimensional challenge it actually is.

I spent weeks piecing together information from dozens of sources, cross-referencing conflicting advice, learning insurance terminology through trial and error, and making costly mistakes. Every question generated three more. Every answer came with unexpected caveats. The process felt deliberately obscure, as if designed to exhaust you before you reached the operating room.

So here’s the guide I wish I’d had—not just tips, but a roadmap through the entire terrain.

A note: I’m not a clinician or attorney. This is lived experience combined with extensive research. Always confirm specific requirements with your insurer, surgeon, and providers.

This isn’t theoretical. I navigated insurance authorization, gathered documentation, found a surgeon who’d actually heard of my procedure, arranged travel from Wisconsin to San Francisco, set up a recovery space, and had surgery—all within nine weeks of my first consultation. It was aggressive. It was exhausting. And it’s absolutely doable if you know what you’re preparing for.

But here’s what I learned: preparation matters more than timeline. Whether you have nine weeks or nine months, the fundamentals stay constant. What changes is your breathing room—how much time you can dedicate to each phase without the pressure of an imminent surgery date. A longer timeline lets you be more methodical, more thorough, more forgiving of inevitable delays. A shorter timeline demands focus, organization, and willingness to make phone calls during lunch and submit paperwork at midnight.

Whether you’re pursuing vaginoplasty or phalloplasty or top surgery or nulloplasty or something the medical literature barely has a name for—the preparation fundamentals are the same. Let me walk you through them.

Define Your Constraints First

Before you dive into research spirals, take thirty minutes and answer four questions. Everything else flows from these:

Insurance or self-pay? This determines your entire financial strategy, which surgeons are accessible, and how much documentation you need.

How far are you willing to travel? Your ideal surgeon may be across the country. Know your travel tolerance before you start falling in love with a practice you can’t realistically get to.

How much time can you take off work? Recovery timelines range from two weeks to six months depending on procedure. If you can’t take unpaid leave, that shapes your options.

Do you have a caregiver available? Most surgeons require a dedicated caregiver for the first one to two weeks. If you don’t have one lined up, that’s a constraint to solve early—not the week before surgery.

Write these down. They’re the guardrails that keep your research focused instead of overwhelming.

Start With Insurance—Right Now

Before you research surgeons, before you daydream about results, before you do anything else: call your insurance company and find out what your plan covers.

This single phone call transforms your entire approach. It determines which surgeons are financially accessible, what documentation you need to gather, and how much time to build into your timeline.

The phone script

Calling insurance while anxious is hard. Here’s exactly what to say:

“Hi, I’m calling to verify coverage for gender affirming surgery under my plan. Can you tell me the specific coverage criteria and prior authorization requirements for [your procedure name]? I’d also like a reference number for this call.”

Then ask:

• Does my plan cover gender affirming surgery according to the WPATH Standards of Care?

• What documentation do I need? (Letters from mental health providers, hormone therapy records, etc.)

• How many referral letters are required, and from what type of provider?

• Is prior authorization required, and how long does it typically take?

Document everything. Date, time, representative’s name, reference number, what they told you. Write it down. I kept a running Google Doc that eventually stretched to forty pages—not because I’m unusually organized, but because insurance companies sometimes “lose” previous conversations. Your own records will save you.

✅ You’re done with this phase when: You have written confirmation of benefits, prior authorization requirements, and clarity on who submits what documentation.

Here’s the frustrating reality: WPATH Standards of Care Version 8 uses criteria that generally involve one referral letter for most genital surgeries, down from the previous requirement of two. But many insurers haven’t caught up. They still require two letters, often with at least one from a doctoral-level provider (Ph.D., Psy.D., M.D., or D.O.). Prepare for the stricter requirements even if your insurer theoretically follows SOC-8. It’s easier to have extra documentation than to scramble for what you need.

If your insurance denies coverage, that’s not the end. Appeals exist, and they work. Some recent federal appellate decisions have found categorical exclusions for transgender surgery discriminatory in certain contexts—but these cases evolve, and the legal landscape continues to shift. Use them as leverage in appeals, not guarantees. Organizations like the Trans Health Project offer free appeal letter templates. Lambda Legal’s Help Desk (866-542-8336) and the Transgender Law Center can help. I’ve seen community members win appeals that seemed hopeless—denials overturned after months of fighting, coverage secured for procedures insurers initially called “cosmetic.”

The system rewards persistence. Treat “no” as the opening bid—ask what documentation would change it, then appeal.

Get Your Letters Early (But Not Too Early)

I thought the letters would be the hardest part. I really did. I built them up in my head as this enormous barrier—finding therapists who understood gender affirming care, scheduling multiple appointments, explaining myself over and over, waiting for providers to write the letters, then waiting for them to be acceptable to insurance and my surgeon. I imagined months of work.

The real barrier wasn’t the letters. It was giving myself permission to get out of my own way first.

Once I stopped treating the process like something I had to earn through extended justification, once I accepted that I knew what I needed and didn’t require anyone’s philosophical approval—just their documentation—the letters happened in a forty-eight hour period. Both of them. From two different providers. Done.

Here’s what actually matters about the letters: Most surgeons and insurance companies require one or two letters from qualified mental health professionals documenting your gender dysphoria and confirming readiness for surgery. The letters need to include a diagnosis of gender dysphoria, confirmation that criteria for surgery are met per applicable standards, confirmation of your capacity for informed consent, confirmation of readiness for surgery and aftercare, the provider’s credentials and contact information, and—this is critical—the specific procedure being requested. Generic letters requesting “gender affirming surgery” get rejected. The letter needs to specify your exact procedure—whether that’s nulloplasty, phalloplasty, vaginoplasty, or any other gender affirming surgery. Be specific.

The timing matters, but not in the way you might think. Letters are typically valid for six to twelve months depending on your surgeon’s requirements and your insurance company’s policies. Get them too early and they expire before your surgery date, becoming expensive paperwork you have to redo. Get them too late and you’re scrambling when you should be focusing on pre-operative logistics. But the window is wide enough that you don’t need to stress about hitting a perfect moment.

✅ You’re done with this phase when: Your letters include the specific procedure name, DSM diagnosis, confirmation of readiness, and the provider’s credentials and contact information.

What I actually needed wasn’t more time with therapists. I’d already done years of therapy. I’d already processed my gender identity, worked through internalized transphobia, built self-acceptance, done the psychological labor that actually matters. What I needed was someone willing to write down what I already knew about myself in the specific format insurance required. That’s a documentation task, not an emotional one.

If cost is a barrier—and therapy appointments add up quickly—GALAP (Gender Affirming Letter Access Project) provides free surgery letters from volunteer mental health providers. This can save you hundreds of dollars while connecting you with providers who affirm rather than pathologize your experience. I’ve heard from multiple people in the community who used GALAP successfully when paying out of pocket wasn’t feasible. It’s a legitimate resource, not a workaround.

The psychological shift that mattered for me was this: I stopped waiting for permission I didn’t actually need. I stopped treating the letters as validation of whether my decision was correct. They’re administrative documentation, not proof of worthiness. Once I understood that distinction, the process moved fast.

Finding the Right Surgeon Is Its Own Full-Time Job

The surgeon who operates on you will shape your body for the rest of your life. That’s not hyperbole—it’s why this decision deserves more rigor than most people give it.

When I started searching, I lived in Southeast Wisconsin. Even Chicago had wait times measured in seasons—not weeks or months, but entire quarters stretching into the next year. When I finally reached a gender clinic at a major university hospital after days of dropped calls and hour-long holds, the earliest consultation was seven months away. Seven months just for an initial conversation about whether I was even a candidate. Worse, the person on the phone had never heard of nulloplasty. She kept redirecting me to vaginoplasty resources, as if I didn’t know what I wanted. As if my request was a misunderstanding rather than a clearly articulated goal.

The experience illustrated a fundamental problem: even in major metropolitan areas with university hospitals and specialized clinics, the infrastructure for gender affirming care remains fragmented, understaffed, and often poorly informed about the full spectrum of procedures people actually seek. If a major research hospital’s intake coordinator doesn’t recognize your procedure, what does that tell you about the depth of expertise available?

Then I remembered a website a friend had shared two years earlier: the Crane Center. When I called them, someone answered within a minute—friendly, knowledgeable, actually helpful. She’d heard of nulloplasty. She knew what I was asking for. She understood the procedure, could speak to the specifics, and didn’t treat my request as unusual or confusing. Days later I had a telehealth consultation. The day after that, I had a surgery date. The contrast between these two experiences—months of gatekeeping versus days of straightforward access—wasn’t about shortcuts or lowered standards. It was about finding a surgical practice that actually specialized in the work I needed done.

Where to search

TransHealthCare.org is the most comprehensive surgery-specific directory currently available, covering 50+ countries and 18 procedure categories. You can filter by insurance type, geographic location, and specific procedures. Treat “featured” listings as advertising, because that’s exactly what they are. Featured placement doesn’t indicate superior surgical outcomes; it indicates a practice paid for visibility. Use the directory as a starting point for generating names, then verify everything independently.

WPATH’s Provider Directory lists thousands of providers globally, but not exclusively surgery-focused. WPATH membership indicates engagement with professional standards and continuing education in transgender health, not necessarily surgical expertise or volume. A provider can be deeply committed to affirming care while performing very few surgeries annually. Verify what each provider actually does, how frequently they do it, and what their specific areas of focus are within the broader field of gender affirming care.

Community wisdom comes from support groups, online forums, social media networks, private conversations with people who’ve been through the process. When someone who’s been through it tells you a surgeon’s results look beautiful but their office never returns calls, that’s information no directory captures. When multiple people mention the same surgeon’s name with genuine enthusiasm or the same practice with consistent warnings, pay attention. Community knowledge exists outside formal channels because it addresses the practical realities of navigating surgical care—the responsiveness of staff, the accuracy of pre-operative information, how complications are handled, whether the surgeon treats you like a person or a procedure.

What to verify

Board certification from a recognized board—ABPS (American Board of Plastic Surgery) for plastic surgeons, ABU (American Board of Urology) for urologists. The American Board of Cosmetic Surgery is not recognized by the American Board of Medical Specialties. This distinction matters. Always ask: certified by which board? Don’t accept vague answers.

Use CertificationMatters.org to verify any board certification claim, and DocInfo.org (provided by the Federation of State Medical Boards) for license status, disciplinary history, and any malpractice actions. These are public databases. Use them. A surgeon unwilling to provide their state medical license number or board certification details is a red flag before you’ve even had a consultation.

Ask about specific volume: a surgeon who’s performed 500 vaginoplasties may have performed only 5 nulloplasties. Volume matters. Technique refinement happens through repetition. Complication management improves with experience. Be specific about your procedure. Ask how many times they’ve performed it in the past year, the past five years, over their career. If they can’t or won’t provide numbers, that’s information too.

Red flags in consultations

Can’t provide specific case volume numbers—”I’ve done many” or “I have extensive experience” aren’t answers. They’re evasions.

Gets defensive when asked about complications—every surgeon experiences complications. It’s not a moral failure; it’s statistical reality. A surgeon who becomes defensive, dismissive, or evasive when you ask about complication rates is telling you they don’t handle difficult conversations well. You need someone who can face complications honestly.

Pressures you to schedule quickly or discourages second opinions—legitimate surgeons want you to make informed decisions. Pressure tactics suggest the practice prioritizes filling their schedule over your wellbeing.

Doesn’t use your correct name and pronouns—if a surgeon can’t get this right in the consultation, they won’t get it right in the operating room, in your medical records, or when complications arise. This isn’t about politeness; it’s about basic respect and competence in providing affirming care.

Promises specific results—surgery involves uncertainty, always. Claims like “you’ll definitely be able to orgasm” or “your sensation will be exactly like before” or “you won’t need revisions” are red flags. Ethical surgeons discuss probabilities, typical outcomes, and the range of possible results—not guarantees.

Green flags

Explains why specific techniques suit your anatomy, not just what they recommend—a surgeon who can articulate their clinical reasoning, who walks you through why they’re suggesting one approach over another based on your specific body, demonstrates depth of expertise.

Discusses alternatives honestly—including the option of not having surgery at all, or waiting, or pursuing a different procedure. A surgeon secure in their expertise doesn’t need to sell you on surgery. They present options and help you make decisions.

Acknowledges complications as a normal part of surgery, not a failure—complications happen. How a surgeon talks about them in consultation predicts how they’ll handle them if they occur in your case.

Treats your choices as valid without requiring justification—you shouldn’t have to defend your gender identity, explain why you want surgery, or justify your specific procedural preferences. A surgeon whose practice centers informed consent respects your autonomy.

That last one mattered to me more than I expected. After years of healthcare experiences where I had to explain and defend my existence, walking into a practice that simply treated my goals as legitimate felt revolutionary. The absence of interrogation, the lack of skepticism, the straightforward acknowledgment that I knew what I wanted—it created space for focusing on medical logistics rather than psychological justification. That’s what affirming care actually means: care that affirms your stated needs rather than questioning whether those needs are real.

Budget for the Costs Nobody Tells You About

Even with insurance, gender affirming surgery involves costs that blindside people. The surgery itself may be covered. Everything surrounding it often isn’t.

Travel: Research published in JAMA found that 56% of vaginoplasty patients and 50% of phalloplasty patients travel out of state for surgery. I traveled from Wisconsin to San Francisco — a distance that added thousands to my total costs but gave me access to a surgeon with specific expertise in the procedure I needed.

Post-operative lodging: Most surgeons require you to stay nearby for 6–14+ days after discharge. At $100–$300 per night, that’s $1,000–$4,000. Some cities have trans-specific recovery housing — Quest House in Oakland, for example, offers rooms at $100–$150 per night.

Caregiver costs: Most surgeons require a dedicated caregiver for the first 1–2 weeks. If they’re traveling with you, budget for their flight, lodging, meals, and lost income. My husband KJ served as my primary caregiver — handling logistics, emotional support, and physical caregiving simultaneously. That level of dedicated support shaped my recovery profoundly, but it came with real financial costs.

Lost income: Minimum recovery times range from 2–4 weeks for orchiectomy to 3–6 months for staged phalloplasty. If you don’t have paid medical leave, this can be the single largest financial impact.

Hair removal: Mandatory for certain surgical sites. At ~$120/hour for electrolysis and requiring 60–100+ hours for genital areas, total costs can reach $7,000–$12,000. Insurance rarely covers it fully. Start this as early as possible — it takes months to years to complete.

Mental health letters: $150–$400 per evaluation session, with most patients needing multiple sessions across one or more providers.

The buffer rule: budget 20–30% above your surgery quote. This handles the surprises without financial crisis.

Grants to apply for:

• Point of Pride Annual Transgender Surgery Fund — the largest resource, with grants ranging from $5,800 to $80,785. Applications open November 1–30 annually.

• DemBois Inc. — grants specifically for trans men of color

• For the Gworls — Black trans-led collective funding rent and surgical costs

• TransMission — $500 micro-grants for various transition expenses

I’ll be honest: for every person who receives a grant, dozens are told no. Apply to everything you’re eligible for. But don’t let your surgery timeline depend on winning.

Prepare Yourself Emotionally (The Part No One Talks About)

Here’s what the clinical literature calls “pre-operative adjustment” and what I call the cruel irony of getting closer to what you need:

Your dysphoria will likely get worse right before surgery resolves it. The closer you get to the finish line, the more unbearable the waiting feels. This is documented, this is common, and this is not a sign that something is wrong with you.

When surgery was a distant dream, you may have protected yourself by not thinking too hard about what you were living with. Once a date is on the calendar, those protective walls come down. The body part you’ve tried not to see in the mirror is all you can see. The dysphoria that was a dull ache becomes sharp and insistent.

And after surgery? The dysphoria doesn’t magically vanish on day one. Swelling obscures results. Healing doesn’t feel like the body you imagined — it feels like a patient’s body. Anesthesia affects brain chemistry for days or weeks. Post-operative depression is common, and reported rates vary widely by study and population — but it’s normal enough that you should expect it, not be blindsided by it.

That’s not a sign of failure. It’s a sign of being human.

What helps:

• Shorten your time horizon. Don’t count the days until surgery. Focus on getting through today.

• Name the paradox: “I feel worse because I’m almost done.” Recognizing the psychological mechanics is validating.

• Know that the body you see at two weeks is not the body you’ll have at two years. Swelling distorts appearance for weeks to months. Scars mature over 12–18 months. Sensation returns gradually over one to two years. Judging outcomes during early recovery causes unnecessary distress.

• Line up mental health support before surgery, not during crisis. A therapist familiar with gender affirming care is worth their weight in gold during the post-operative emotional rollercoaster.

• Let yourself grieve and celebrate simultaneously. You can be excited about what’s coming and heartbroken that it took this long. You can be scared and certain. Human emotional life is not a coherent narrative; it’s a swirl. Let it swirl.

The long-term data is reassuring: satisfaction rates for gender affirming surgery exceed 90% across procedures, regret rates average approximately 1%, and a 40-year follow-up study found that benefits persist for decades. Between the operating room and those long-term outcomes lies a hard stretch. But you get through it.

The Timeline: What to Do When

Adjust based on your own situation, but this is the skeleton:

12+ months out Start electrolysis/laser if your procedure requires it. Begin researching surgeons and insurance coverage.

6–12 months out Call insurance, verify coverage, get it in writing. Begin gathering mental health letters. Schedule surgical consultations.

3–6 months out Submit prior authorization through your surgeon’s office. Complete pre-op requirements (labs, medical clearance, procedure-specific prep). Begin financial planning for hidden costs.

6–9 weeks out Finalize travel and lodging. Set up your recovery space. Stock supplies. Arrange your caregiver. File FMLA paperwork if applicable. Fill prescriptions.

1–2 weeks out Confirm surgery date, arrival time, and last-minute instructions. Pack your hospital bag. Download entertainment. Tell your support network your timeline.

The day before Follow pre-op instructions exactly. Eat well (before the fasting window). Lay out everything you need. Try to sleep — and forgive yourself if you can’t.

The Truth About Preparation

Here’s what I’ve learned from my own surgical journey and from supporting dozens of others through the Flat Crotch Collective and bottomsurgery.info:

The people who navigate this process successfully tend to start early. Not a month or two before surgery, but as far out as they can. Insurance authorization alone can take months. Legal name changes vary from weeks to half a year. The more lead time you give yourself, the less stress you’ll experience when the inevitable delays hit.

Documentation is your armor. Every phone call logged, every email saved, every letter filed. This becomes invaluable when you need to appeal a denial, prove what you were told, or simply remember where you are in a process that stretches across months.

Building a team matters. Your surgeon’s insurance coordinator, a knowledgeable therapist, a supportive employer, online community members who’ve been through it, and potentially legal advocates. You don’t need everyone in place before you start. But knowing who can help — and being willing to ask — makes the journey more manageable.

Flexibility is harder than it sounds, but essential. Plans will change. Dates will shift. Denials will require appeals. My own surgery date moved three times before it finally happened. Each time felt devastating in the moment. Looking back, none of those delays changed the outcome — they just tested my patience.

And please: take care of yourself through the process, even when the process fights back. Pace yourself. Take breaks. Celebrate small victories — getting that call returned, receiving that document, finally understanding what your policy actually says. The paperwork is a means to an end, not the end itself.

You’ve been waiting long enough. Let’s get you ready.

This article draws from my book Becoming Yourself: The Complete Guide to Gender Affirming Surgery — 40 chapters covering every procedure, every country, every system you need to navigate. It’s the book I needed and it didn’t exist, so I wrote it. Available now on Gumroad.

I also maintain bottomsurgery.info as a free resource hub, and run the Flat Crotch Collective — a 100+ member community for people pursuing or curious about nulloplasty.

Questions about your own surgical prep? Drop them in the comments or reply to this email. I read everything.

2026 is here, and with it comes a shift I’ve been building toward for longer than I can remember. I’m done compartmentalizing my digital life. And I’m done trying to cram myself into structures that were never built for someone like me.

For years, I’ve split myself across platforms and personas like some kind of fractured mirror. There was the “professional” version of me: the one posting code, diving into tech projects, sharing thoughtful insights on systems and structures. Clean, focused, palatable. Then there was the private me, the one exploring kinks, raw desires, unfiltered sexuality, and the messy edges of being human. Separate accounts, careful boundaries, endless mental energy spent curating what parts of me were allowed to breathe in public. It kept everything “safe,” or so I told myself.

But it was exhausting. It drained me in ways I didn’t fully register until recently. Living like that meant constantly performing, shrinking pieces of myself to fit whatever audience I thought I was facing. That sanitized, perfect version? The one who never got too weird, too horny, too real? It never actually existed. It was a ghost I was chasing, a facade that kept me from ever feeling whole.

The same was true professionally. I spent years trying to fit into corporate structures, convincing myself that if I just found the right company, the right role, the right team, it would finally click. It never did. The truth I kept dodging is simple: I’m not built for that. I’m an entrepreneur. I’m a creator. I need to build things my way, on my terms, without asking permission or filtering myself through someone else’s org chart.

And here’s the thing: I have the credentials. A Finance MBA from Johnson & Wales. A BA in Political Science, History, and International Studies from UW-Madison, where I graduated Phi Beta Kappa. An AAS in Culinary Arts, because my path has never been linear and I refuse to pretend otherwise. I’ve spent years accumulating skills across wildly different domains, and now I’m finally ready to deploy them on my own terms. It’s time to believe in and seize my own power and place in the world.

So I’m retiring all of it. The compartmentalization. The corporate aspirations. The careful curation. Starting now, all of my accounts are going to integrate the full unfiltered spectrum. My code and my porn will live side by side, no walls between them. Serious deep dives into projects will sit right next to horny late-night thoughts. Celebrations of wins will mingle with the weird, vulnerable, joyful chaos of being alive in this body, this mind, this life.

No more fragments. No more hiding parts of myself to make others comfortable.

And I’m not just integrating my identity. I’m integrating my entire creative output across every platform I touch. Merch on Threadless. Porn on JustForFans. Code and porn and everything in between on X and Bluesky. Long-form writing on Substack. All of it connected, all of it me, all of it building toward something that’s mine. I finished recovering from my revision surgery in November, and now I’m ready to build. Really build. Not as an employee hoping for a seat at someone else’s table, but as an independent creator constructing my own nullo creative empire.

While we’re on the topic of things I’m not apologizing for: I’m a proud user of generative AI in my work. I use it across my entire stack, everything I do, every single day. I have teams of agents continually doing work for me in the background.

What does “AI workflow architect” actually mean? It means I’ve built systems where AI isn’t a novelty or a shortcut—it’s infrastructure. Claude for reasoning, writing, research, and decision-making. Linear as a persistent memory layer where AI agents track, update, and execute tasks autonomously—not a human project management tool, but an AI-native one. Notion as my knowledge base and long-term storage,m with Factory.ai orchestrating automations across all of it. MCP servers connecting everything so the pieces actually talk to each other.

I treat these tools like employees with specific roles, not chatbots I poke when I’m bored. They have context. They remember. They execute while I sleep. I’ve spent months building, testing, and refining these workflows—figuring out what works, what breaks, what scales. This isn’t prompt engineering. This is systems design.

I wrote a 1000+ page book in months, not years. I run a community, manage my own healthcare bureaucracy nightmares, create content across platforms, and still have energy left over. That’s not because I’m superhuman. It’s because I’ve built a machine that multiplies what one person can do.

And I’m done pretending the traditional job market makes any sense anymore. The whole thing has become Black Mirror absurdist theater—AI screening out human applicants, ghost jobs that don’t exist, hiring managers who never respond, systems designed to make you feel lucky to be exploited. I’m not playing that game. I’m building my own thing, on my own terms, with tools that actually work for me instead of against me.

If any of that offends you, I’d ask you to really examine why you’re so concerned about how other people do their work. Focus on your own.

If you followed me expecting only one narrow slice (my ass), you might be surprised by what shows up now. That’s okay. If the full human resonates, stick around; I’d love to have you here. If it doesn’t, unfollows are free and come with zero hard feelings. This isn’t about chasing numbers or approval. It’s about finally showing up as one complete person, building one complete body of work.

This feels like the truest way to step into actually becoming myself: loudly, messily, completely, without apology.

Which brings me to the first major release from this newly integrated creative life. I just published my book: Becoming Yourself: The Complete Guide to Gender-Affirming Surgery.

This isn’t a short read or a surface-level overview. It’s a 1000+ page, 40-chapter comprehensive guide, densely packed with rigorously researched information, written entirely from the perspective of a patient who lived it. I underwent nulloplasty in August 2024 at The Crane Center. Throughout that entire arc, from the first flickers of decision-making to long-term thriving on the other side, I kept wishing for a resource that didn’t exist: something honest, tactical, and complete, written by someone who had actually been on the table, not just standing beside it.

Most of what’s out there falls short in one way or another. Clinical literature gives you third-person statistics and detached overviews. Useful data, but none of the lived feeling of what it’s actually like to navigate dysphoria, recovery, or the emotional rollercoaster. Memoirs tell beautiful personal stories, but they often prioritize narrative over step-by-step guidance. Online forums like Reddit and Discord are treasure troves of real experiences, yet they’re scattered, frequently outdated, full of conflicting advice, and hard to verify. And when it comes to complications (the scary, real ones that can derail everything), most sources either gloss over them or present them in cold clinical terms that don’t prepare you for the human reality.

This book is different. It’s the first comprehensive guide written entirely by a patient, and the only one with truly in-depth coverage of nulloplasty alongside the full range of gender-affirming procedures. It weaves my first-person experience throughout every section, but never at the expense of actionable tools. It’s structured to take you through the entire journey with clarity and honesty.

The stakes in this process are enormous, and that’s why preparation matters so much. Choose the wrong surgeon and you could be facing a $30,000 revision, or worse. Miss one obscure insurance requirement and your claim gets denied, derailing months or years of progress. Skip key prep steps and you risk preventable complications that could have been avoided. And without a trusted central resource, you end up wasting dozens (sometimes hundreds) of hours sifting through noise, trying to separate signal from misinformation.

This guide exists to eliminate that chaos. It’s built as a tactical playbook for anyone, anywhere in the world, pursuing any form of gender-affirming surgery. Whether you’re just starting to explore options or you’re deep in the logistics, it turns overwhelm into a clear path forward.

The first volume walks you through preparation, procedures, and recovery in exhaustive detail. It begins with the foundational work: truly understanding dysphoria, debunking persistent myths, and reframing how we think about bodies and identity. From there it moves into the practical logistics. HRT foundations, navigating insurance systems, handling legal documentation, fertility preservation, considerations for neurodivergent folks, and how to find the right surgeon for you. The procedure chapters go deep: every major technique, realistic timelines, risks that actually matter, and all the “real considerations” that clinical pamphlets conveniently omit. Recovery gets the attention it deserves, including managing dysphoria in those vulnerable post-op weeks. There’s also an entire dedicated caregiver guide for partners, parents, or chosen family, with practical advice on avoiding burnout, handling crises, and offering meaningful emotional support.

The second volume looks beyond the operating room to life afterward, and to access challenges around the globe. It covers physical recovery in the long term, rebuilding intimacy and relationships, sex after surgery, and what thriving really looks like when the dust settles. For those seeking care outside their home country, there are detailed frameworks and country-specific guides for twelve different locations, helping you navigate international options with eyes wide open.

Scattered throughout are seventeen practical appendices: tools designed to be used, not just read. Things like a 9-week surgery sprint plan to keep you on track, a caregiver quick-reference card for emergencies, an international packing list, insurance appeal templates, and trackers for medications and recovery milestones.

One of the highest-value pieces so many people have already told me saved them headaches: eight fully SOC-8 compliant WPATH therapy letter templates. Therapists often want to help but struggle with the exact clinical language insurers demand. These templates provide it word-for-word, structured to get approvals on the first try instead of facing delays and rejections.

This book is written first and foremost for trans and nonbinary people who are considering or actively preparing for surgery, especially those exploring nulloplasty, where resources have been particularly thin. But it’s also for the people who love and support them: partners and spouses who need to understand the timeline and emotional weight, caregivers looking for concrete ways to show up, therapists seeking deeper client insight and ready-to-use tools. And it’s for allies: friends, family members, coworkers, anyone who cares about the trans people in their lives and wants to truly understand what they’re going through. Not surface-level allyship, but real comprehension of the fears, the logistics, the recovery, the joy. If you’ve ever wanted to support someone through this process but didn’t know how to ask or what to say, this book will give you that understanding.

If any part of this journey has ever felt lonely or overwhelming, this guide is meant to feel like a companion that’s already walked the path and left detailed notes for you.

It’s available now as a digital PDF/e-book for $49 USD on Gumroad!

If you’re not ready to commit to a $49 bottom surgery encyclopedia, just check out https://bottomsurgery.info. I’m producing abundant free resources there, and I’m planning a massive site expansion now that my 1000+ page beast is finally out of me. It won’t have nearly the scope of the book (because of the hundreds of hours I spent producing that information), but it will give you a place to start. You can also follow along here on Substack!

Subscribe now

Lastly, if my path resonates with you and you need community support to help you become yourself, consider applying to join the Flat Crotch Collective. We’re a 100+ strong Telegram community supporting each other through everything: surgery, mental health, the attacks on the trans community, all of it. This isn’t a fetish group or a fan club. It’s for people who know, or at least highly suspect, that this path is for them. You can apply here.

Thank you for being here as I step into this new chapter, both personally and professionally. 2026 feels wide open. Here’s to showing up fully, building freely, becoming ourselves without apology, and supporting each other along the way.

~ Axl

If you’re reading this and noticing a new name at the top, yep.

This newsletter is now Flat Crotch Dispatch.

That name is not me trying to be clever. It is me being accurate. This is where I write from inside gender-affirming surgery and the life that comes after it. The logistics nobody lays out cleanly. The recovery truths that don’t fit in a clinic handout. The stuff you usually have to learn the hard way.

I’m renaming it now because two things became true at the same time: I decided to stop splitting myself into different versions online, and I finally finished my book.

The book is titled Becoming Yourself. While writing 900+ pages about clarity, decision-making, and identity, I realized I couldn’t keep living behind a wall of separated names. So I’m taking my own advice. I’m showing up as the same person everywhere: Axl Ibiza / @borgpup.

And the book isn’t a draft anymore. It’s done!

Becoming Yourself: The Complete Guide to Gender Affirming Surgery

It is available right now for $49.

Here is exactly why it earns that price.

I wrote the book I couldn’t find

When I was preparing for nulloplasty, I searched extensively for a comprehensive guide that would walk me through the entire process from start to finish. Instead of finding that cohesive resource, what I discovered was the modern trans-health scavenger hunt: Reddit threads filled with outdated information and dead links, scattered blog posts across the internet that were only partially correct or incomplete, and clinical brochures that were clearly written for medical professionals rather than for patients who needed practical, actionable guidance.

The comprehensive resource I desperately needed simply didn’t exist anywhere. So I made the decision to spend two full years researching, documenting, and creating it myself.

The result is 40 detailed chapters. 1,093 pages of thoroughly researched content. Information gathered from experiences across 12 different countries.

It covers the complete journey from beginning to end: Decision-making → insurance navigation → surgeon selection → preparation protocols → surgery itself → recovery process → handling complications → long-term thriving and sustainability.

One comprehensive resource. Everything you need in a single place.

The only book of its kind

I’m not going to hedge this statement or soften it in any way. This is the only book about nulloplasty that has ever been written and published.

Not merely the first book on this subject. The only one that currently exists.

Beyond that distinction, it is also the very first comprehensive, patient-written guide to gender-affirming bottom surgery that actually functions like a practical reference tool you can use throughout your journey. Everything else that’s currently available out there is either a personal memoir focused on individual experience or a dense medical textbook written for clinical professionals. This is a tactical manual written by someone who has actually been on the operating table and lived through the entire experience firsthand.

Surgeons can tell you in great detail what they do during the procedure, but they fundamentally cannot tell you what it’s like to actually be the person being operated on. They can explain the surgical technique with precision, but they can’t tell you what the recovery process feels like day by day, how to manage post-operative dysphoria when it appears, or what happens when complications hit and you’re the one living inside your body dealing with the consequences in real time.

This book documents it all—including the serious complications and difficult moments—from the inside perspective of someone who has been through it. Not to scare you or discourage you from pursuing surgery, but so you can make informed decisions with your eyes fully open to both the possibilities and the realities.

The stakes are too high for bad information

At this current moment in time, the process of preparing for gender-affirming surgery typically involves collecting and assembling scattered fragments of information from dozens upon dozens of different sources spread across the internet, and then hoping against hope that you haven’t accidentally overlooked or missed something critically important along the way. But the potential consequences and stakes involved here aren’t merely “oops, I accidentally wasted an afternoon of my time.”

Gender-affirming surgery represents a financial investment ranging anywhere from $20,000 to well over $150,000+ out of pocket depending on the procedure and location. It is a permanent, irreversible change to your body. It is fundamentally life-changing in ways that will affect you for the rest of your existence.

Making one wrong choice when selecting your surgeon can result in a $30,000 revision surgery.

Missing or overlooking one single insurance requirement or documentation step can lead to denied coverage and having to pay out of pocket.

Skipping or not knowing about one crucial preparation step can result in preventable complications that could have been avoided.

You deserve to have access to complete, thoroughly organized, and genuinely reliable information that you can trust as you navigate this journey.

What is actually inside (The Receipts)

This is intentionally structured and organized as two distinct volumes, so it’s not simply an overwhelming mass of pages thrown together—it’s a deliberately designed, systematically organized, and highly usable reference system that you can actually navigate and apply to your specific situation.

Volume 1 (Preparation & Recovery) covers the complete journey and entire arc from start to finish: comprehensive HRT foundations and hormone therapy considerations, detailed legal documentation requirements and processes, neurodivergence and how it intersects with surgical preparation and recovery, practical guidance for finding and vetting a qualified surgeon, and extensive deep dives into every major surgical procedure with detailed explanations of what each entails. Crucially and importantly, it includes a Comprehensive Caregiver Guide that’s specifically written for the support people in your life, because the people who are taking care of us throughout this process deserve to have their own detailed playbook and resource guide too.

Volume 2 (Long-Term & International) covers everything that comes after surgery and life after the bandages finally come off: detailed physical recovery protocols and timelines, navigating relationships and intimacy after surgery, managing your mental and emotional health during the long recovery process, and a massive, comprehensive framework for international care with detailed country-specific guides that break down the unique considerations, costs, legal requirements, and logistics for pursuing surgery outside your home country.

The “I Need It Now” Section

The book includes 17 highly practical and immediately actionable appendices designed to get you moving forward immediately without delay, including detailed surgeon interview questions you should ask before committing, a structured 9-week surgery sprint plan for when your date gets scheduled, comprehensive recovery trackers to monitor your healing progress, and checklists to ensure you don’t overlook critical preparation steps.

Plus: 8 WPATH Therapy Letter Templates

For many people navigating the insurance approval process, securing the required therapy letters becomes an expensive and time-consuming bottleneck that delays everything. Many therapists and mental health providers are genuinely willing and eager to write these letters for their patients, but they frequently get rejected by insurance companies because they don’t know the highly specific language, formatting requirements, and documentation standards that insurers require and expect to see. These templates are SOC-8 compliant frameworks and structured guides that your provider can directly reference and adapt to get the letter written correctly and accepted the first time. Stop wasting weeks or even months going back and forth on formatting errors and missing documentation requirements.

This should be required reading for every healthcare professional

I’m going to make a bold claim here, and I mean it without qualification: Every single healthcare professional who interacts with trans patients—surgeons, nurses, therapists, primary care physicians, insurance coordinators, social workers—should be required to read this book.

Not as a suggestion. Not as “recommended professional development.” As a requirement.

Here’s why:

Healthcare providers consistently overestimate what they know about the patient experience. They know their procedures. They know their protocols. They know what happens in the operating room and what the post-op instructions say. What they fundamentally do not know—and cannot know without resources like this—is what it actually feels like to navigate insurance denials at 2 AM, what complications look like from the inside when you’re alone in your bathroom, or how to make decisions when every online source contradicts the last.

The information gap kills trust. When a surgeon tells a patient “recovery is straightforward” and that patient then experiences complications that were never mentioned, that’s not just a communication failure—it’s a betrayal. When a therapist writes a letter that gets rejected because they didn’t know the SOC-8 requirements, that’s not the patient’s fault—it’s a systems failure that could have been prevented with proper training and resources.

This book documents what your patients are actually living through. The financial terror. The research paralysis. The surgical preparation nobody explains properly. The recovery realities that don’t match the brochure. The long-term adjustments that extend far beyond the 6-week follow-up. If you are providing care to trans patients pursuing bottom surgery and you have not read a comprehensive patient perspective on this process, you are operating with incomplete information.

Patient-written resources make you a better provider. This isn’t about replacing medical education—it’s about supplementing it with the perspectives that medical training systematically excludes. The best healthcare professionals I encountered during my journey were the ones who actively sought out patient narratives, who asked what resources I was using, who acknowledged the limits of their own perspective and worked to understand mine.

Reading this book won’t make you an expert in the patient experience. But it will make you a more informed, more empathetic, and more effective provider. It will help you ask better questions. It will help you spot gaps in your own communication. It will help you understand why your patients show up to appointments with 47 browser tabs open and a spreadsheet of questions.

If you care for trans patients, you owe it to them—and to yourself—to understand what they’re actually navigating. This book is that understanding, documented across 1,093 pages of lived experience and comprehensive research.

Make it required reading. Make it part of training protocols. Make it something every provider in the gender-affirming care space has actually engaged with, not just passively recommended.

Your patients are already doing the work of educating themselves. Meet them halfway.

The only comprehensive guide written for caregivers, not just patients

Here’s something that makes this book genuinely unprecedented in the gender-affirming surgery space: it includes a complete, detailed Caregiver Guide specifically written for spouses, partners, family members, and primary support people.

This isn’t a brief “tips for helpers” sidebar tucked into a chapter. It’s a comprehensive, standalone resource that treats caregivers as full participants in this journey—because that’s exactly what they are.

Every other resource about gender-affirming surgery is written exclusively for the patient. And while patients obviously need detailed information and guidance, the reality is that surgery and recovery fundamentally cannot be done alone. You need people. You need support. You need someone to drive you home from the hospital, manage medications when you’re too foggy to remember, help you to the bathroom when mobility is limited, and hold space for the emotional complexity that emerges during recovery.

But here’s the problem that nobody talks about: caregivers are expected to show up and provide complex medical and emotional support with almost zero preparation, training, or guidance.

They don’t get the pre-op consultations. They don’t get the detailed recovery protocols. They don’t get the surgeon’s explanations. They’re handed a patient who just underwent major surgery, given a few basic instructions, and expected to figure out the rest on their own.

That’s not fair to them. And it’s not safe for patients.

The Caregiver Guide in this book changes that entirely.

It walks support people through everything they need to know: what to expect during each phase of recovery, how to recognize warning signs and complications, how to provide practical physical care without causing harm, how to support someone emotionally when they’re navigating post-surgical dysphoria or unexpected challenges, and crucially—how to take care of themselves so they don’t burn out in the process.

It acknowledges that caregivers have their own fears, questions, and emotional responses to this process. It gives them language for conversations they don’t know how to start. It prepares them for realities that patients themselves might not be ready to articulate yet.

This matters immensely for relationship sustainability. Surgery doesn’t just change the patient’s body—it changes the entire dynamic of intimate partnerships and family systems. When caregivers feel equipped, informed, and prepared for what’s coming, they can show up more effectively. When they’re flying blind and improvising everything, resentment builds, communication breaks down, and relationships suffer under the strain.

I wrote this Caregiver Guide because I watched my own support people struggle with questions they didn’t know how to ask. I saw them second-guessing their decisions, wondering if they were helping or hurting, carrying anxiety they had no outlet for because all the resources were focused exclusively on me.

They deserved better. Your support people deserve better.

If you’re going through this process, buying this book isn’t just an investment in your own preparation—it’s a gift to the people who will be taking care of you. Give them the comprehensive guide they won’t find anywhere else. Let them show up informed, confident, and genuinely prepared for what’s ahead.

And if you’re a caregiver reading this: you don’t have to figure this out alone. The resource you need exists now. It was written for you.

This book is for anyone who gives a damn about building a better world

Let me be absolutely clear about something: You do not have to be trans to read this book.

You do not need to be planning surgery. You do not need to be questioning your gender. You do not need to have a trans family member or close friend.

If you want to be a better ally to a community under constant attack and duress, read this book.

Here’s why that matters:

Allyship without understanding is performative at best, harmful at worst. Posting a pride flag once a year doesn’t cut it. Saying “I support trans people” while having zero comprehension of what trans people actually navigate, endure, and survive—that’s not solidarity. That’s decoration.

This book will show you, in unflinching detail, what trans people are actually up against. The financial barriers that make life-saving care inaccessible. The insurance systems designed to deny and delay. The medical gatekeeping that treats our autonomy as negotiable. The information deserts that force people to become their own researchers, advocates, and experts just to survive.

Reading this book will make you angry. Good. That anger is appropriate. Channel it into action that actually matters.

Understanding the reality makes you a more effective advocate. When you know what the insurance denial process actually looks like, you can push back on policy in informed ways. When you understand what surgical recovery actually entails, you can offer meaningful support instead of empty platitudes. When you grasp the financial burden, you can advocate for systemic change that addresses root causes.

This isn’t about centering yourself in someone else’s story. It’s about doing the work to understand what you’re claiming to support. It’s about recognizing that trans liberation is bound up with everyone’s liberation—that the systems oppressing trans people are the same systems harming everyone who doesn’t fit neatly into rigid boxes.

If you care about bodily autonomy, read this book.

If you care about healthcare justice, read this book.

If you care about dismantling systems that gatekeep access to necessary care, read this book.

The trans community is under relentless legislative attack. Our healthcare is being criminalized. Our existence is being debated in legislatures and school boards and dinner tables. People are making decisions about our lives without having the faintest idea what our lives actually look like.

You can choose to be different. You can choose to actually understand before you speak. You can choose to learn what this community faces before you claim to stand with us.

This book is 1,093 pages of that education. It’s not easy reading. It’s not comfortable. It will challenge assumptions you didn’t know you had.

Do it anyway.

Because if you genuinely want a better world—not a theoretical better world, but one where trans people can access necessary healthcare without jumping through dehumanizing hoops, where we’re not bankrupted by the cost of survival, where we’re treated as the experts on our own bodies and lives—then you need to understand what we’re fighting against.

And you need to fight alongside us with actual knowledge, not just good intentions.

Read this book. Then use what you learn to be a better ally.

We need people who understand. We need people who will show up informed and effective when it matters most.

Be one of those people.

Get the book

Becoming Yourself: The Complete Guide to Gender Affirming Surgery

One comprehensive resource. Everything you need to navigate this journey successfully.

Written by someone who’s actually been through the entire experience firsthand.

Welcome to Flat Crotch Dispatch. This newsletter is the field notes version of the work—the ongoing, real-time documentation of life inside gender-affirming surgery and recovery. If you’ve been waiting and searching for the comprehensive resource that simply didn’t exist anywhere:

It exists now. You can access it today.

Axl Ibiza / @borgpup

PS: If you know someone who is currently preparing for surgery and is drowning in conflicting advice, scattered information, and overwhelming uncertainty, please forward them this post. They don’t have to navigate this journey alone or figure everything out by themselves.

Share Flat Crotch Dispatch

Subscribe now

If you’re reading this, maybe you probably haven’t fully committed to AI tools yet. Maybe you’re skeptical. Maybe you’re concerned about the ethics. Maybe you’re waiting to see how things shake out. Maybe you’re managing just fine with the systems you already have.

I understand. I also need you to understand what that waiting may be costing you—and who can actually afford to wait.

But this isn’t just another article telling you you’re “behind.” The conversation about AI has become frustratingly binary: breathless enthusiasm on one side, principled rejection on the other. Neither extreme serves most people. What we need is a more honest conversation about what these tools can actually do, who they help most, what the genuine risks are, and how to adopt them thoughtfully rather than frantically.

Rejecting AI tools on moral grounds has become a marker of intellectual sophistication in certain circles. But this position is increasingly untenable for most people. The productivity gap between AI adopters and non-adopters is already measurable and widening. For disabled, neurodivergent, and resource-constrained individuals, this isn’t a philosophical debate about technology’s soul—it’s a question of whether they can participate at all.

The uncomfortable truth is that AI rejection discourse often emanates from positions of extraordinary privilege: tenured academics invoking academic freedom, established professionals with secure positions, and well-funded organizations that can absorb inefficiency. Meanwhile, those who most need productivity tools to level the playing field—the freelancer with ADHD, the researcher with chronic illness, the entrepreneur competing against venture-backed companies, the aging parent trying to manage increasingly complex healthcare—cannot afford the luxury of principled abstention.

But I need to be honest from the start: adopting AI is also a privilege. And the tools I’m advocating for have real limitations, real risks, and real costs that deserve serious engagement rather than dismissal. This piece attempts that engagement.

Part I: The Evidence

The productivity gap is already here

Let me show you the numbers. They’re not projections—they’re measurements from controlled studies.

A landmark MIT and Stanford study tracked 5,179 customer service agents using AI tools (Brynjolfsson et al., 2023). The average productivity increase was 14% (measured as issues resolved per hour). But the effects were dramatically uneven. Newer and lower-skilled workers saw 34% productivity gains. Workers with two months of tenure plus AI performed as well as workers with six months of tenure without it. The time to reach proficiency was cut in half. Important context: “productivity” here is specific to customer service metrics, not a universal guarantee for all knowledge work.

Read that again: two months with AI equals six months without.

A Harvard Business School study of 758 consultants at Boston Consulting Group found even starker disparities (Dell’Acqua et al., 2023). Lower performers improved by 43%. Top performers improved by 17%. Everyone completed 12.2% more tasks, 25.1% faster, with 40% higher quality outputs. AI didn’t just help—it compressed the skill gap.

GitHub’s controlled experiment with 95 developers found that those using Copilot completed coding tasks 55.8% faster (Peng et al., 2023). Not marginally faster. More than half again as fast.

And these gains compound. OpenAI’s enterprise productivity report found that “frontier workers”—those who have deeply integrated AI into their workflows—send six times more messages to AI systems than the median user and save more than ten hours per week (OpenAI, 2025). The gap in coding tasks is even more extreme: a 17-to-1 ratio between power users and average users.

The BCG 2025 report on AI adoption found that companies that have built AI capabilities see labor productivity grow 1.7 times faster than the global average and generate 3.6 times higher total shareholder returns (Boston Consulting Group, 2025). McKinsey reports that organizational AI adoption has risen from 55% in 2023 to 72% in 2024 to 78% in 2025, with two-thirds of organizations expecting to increase AI investment further (McKinsey & Company, 2024, 2025).

Part II: Disability and Neurodivergence

Disability is the statistical norm, not the exception

The framing of AI as “optional” for “normal” people rests on a fundamental misconception: that disability is rare and exceptional. The data tells a radically different story.

According to the Social Security Administration’s actuarial tables, women have a 54% chance of becoming seriously disabled during their working years; men face 43% odds (Social Security Administration, n.d.). One in four 20-year-olds entering the workforce today will be out of work for at least a year due to disability before retirement. Almost 90% of long-term disabilities are caused by illness, not accidents—meaning virtually anyone can join this category without warning (National Safety Council, 2008).

The numbers compound with age. By 65, approximately 85% of Americans have at least one chronic health condition. The “non-disabled” category is not a stable identity but a temporary condition—what disability scholars call “temporarily able-bodied.”

Add neurodivergence to this picture and the notion of a “normal” baseline becomes even more untenable. Research published in the British Medical Bulletin estimates that 15–20% of the global population is neurodivergent (Doyle, 2020). This encompasses ADHD, which a recent study found affects 11.4% of children (Danielson et al., 2024); autism spectrum conditions, now identified in 3.2% of children according to the latest CDC surveillance (Shaw et al., 2025); and dyslexia, which the Yale Center for Dyslexia & Creativity estimates affects up to 20% of the population. These conditions frequently co-occur.

If you include everyone who is currently disabled, will become disabled, or is neurodivergent, you’re no longer talking about an edge case. You’re talking about most people at some point in their lives.

AI as assistive technology

For neurodivergent individuals, AI tools function as cognitive prosthetics—external scaffolding for executive function, attention management, and task organization that neurotypical individuals take for granted as internal capabilities.

A 2024 Ernst & Young study of 317 neurodivergent and disabled employees across 17 organizations found that 76% report AI tools help them thrive more at work, while 85% believe these tools can create more inclusive workplaces (Ernst & Young LLP, 2024). It is important to note that this study was conducted “in collaboration with Microsoft” to evaluate Microsoft’s Copilot product; industry-funded research on industry products warrants appropriate scrutiny. That said, the study found that 91% see AI assistants like Microsoft Copilot as helpful assistive technology, and 88% feel more productive when using them.

The mechanisms are specific and measurable. For individuals with ADHD, AI assistants serve as external regulators for attention management, breaking down complex tasks into manageable steps, providing adaptive support, and reducing cognitive load. For individuals with dyspraxia, AI reduces stress and anxiety in written communication. For those with speech and writing disabilities, AI provides robust communication support previously unavailable at any price point.

The parallel to physical assistive technology is instructive. Screen readers, voice interfaces, speech-to-text systems, and augmentative communication devices have transformed participation for people with sensory and physical disabilities. AI now enhances all of these: Be My Eyes combined with GPT-4o provides instant image descriptions for blind users; Google Live Transcribe offers real-time speech transcription for deaf individuals; AI-powered augmentative communication devices give voice to non-speaking people.

Consider Congresswoman Jennifer Wexton, whose progressive supranuclear palsy took her ability to speak. Using ElevenLabs’ AI voice cloning technology, she became the first member of Congress to speak on the House floor using an AI-generated voice (Sprunt, 2024; Wexton, 2024). The technology didn’t replace her—it restored her capacity to participate.

These aren’t luxuries. They’re infrastructure for basic participation.

The employment gap these tools could close

The employment disparities for disabled and neurodivergent people are staggering. According to the Bureau of Labor Statistics, only 22.7% of disabled individuals ages 16–64 are employed, compared to 65.5% of non-disabled Americans (U.S. Bureau of Labor Statistics, 2025).

For autistic adults, the situation is worse. The National Autistic Society found that only 16% of autistic adults in the UK are in full-time paid work (National Autistic Society, 2016). UK Office for National Statistics data shows that 21.7% of autistic people are employed—the lowest rate of any disabled group (Office for National Statistics, 2020). The widely cited statistic that 85% of autistic adults are unemployed or underemployed comes from aggregating these studies with U.S. data from Drexel University’s autism outcomes research (Roux et al., 2015).

AI tools that could help close these gaps become inaccessible when rejection is moralized. When we tell people they should accomplish work “without AI,” we’re saying something structurally identical to telling disabled people they should accomplish tasks without wheelchairs, screen readers, or medication. We’re establishing a standard that systematically excludes those who need tools to participate.

Part III: The Adoption Paradox

Who can afford to say yes?

I’ve argued that rejecting AI is a privilege. But intellectual honesty requires acknowledging the inverse: adopting AI is also a privilege.

The freelancer with ADHD I’ve invoked as needing these tools? Before AI can help her, she needs:

• Digital literacy: Comfort with typing, navigating interfaces, evaluating output

• Hardware: A computer or smartphone capable of running modern applications

• Connectivity: Reliable internet access (still unavailable to 24 million Americans)

• Financial resources: $20/month for Claude Pro or ChatGPT Plus—$240/year, roughly equivalent to a week’s groceries for a family on SNAP benefits

• Time and cognitive bandwidth: The executive function to learn new tools before they help with executive function

• Critical evaluation skills: The ability to recognize when AI output is wrong, incomplete, or hallucinated

This is the adoption paradox: the populations who most need AI’s assistive capabilities often face the highest barriers to accessing them. The 72-year-old overwhelmed by medical paperwork may also be overwhelmed by the prospect of learning yet another digital system. The person with cognitive decline who could benefit from memory assistance may lack the cognitive resources to set up that assistance while they still can.

This doesn’t negate my argument—it complicates it. The question isn’t simply “reject or adopt?” but “how do we ensure the people who most need these tools can actually access them?”

Several implications follow:

The paper-to-digital gap is the real barrier. For millions of people, the obstacle isn’t AI skepticism—it’s that they haven’t crossed the digital divide that AI adoption presupposes. Advocacy for AI adoption that ignores this gap is advocacy for the already-privileged.

Cost matters more than tech discourse acknowledges. Twenty dollars a month is trivial for a knowledge worker; it’s a real tradeoff for someone on fixed income. Free tiers exist but offer degraded capabilities. The productivity gap I’ve documented may become a wealth gap if access isn’t democratized.

Learning curves impose cognitive taxes. The promise that AI reduces cognitive load assumes you’ve already paid the cognitive cost of learning to use it. For someone already overwhelmed, that upfront investment may be prohibitive—even if the long-term payoff would be substantial.

Support systems matter. The most successful AI adoption I’ve observed happens when someone with digital fluency helps someone without it. Adult children helping aging parents. IT departments supporting employees. Community organizations bridging gaps. Atomized individual adoption isn’t enough.

This reframes the moral stakes. The problem isn’t just that privileged people moralize against AI adoption while others can’t afford to abstain. It’s that both positions—reflexive rejection and breathless adoption—center the experiences of the digitally fluent and economically secure. The person managing life on paper systems, the elder without family tech support, the worker without employer-provided tools—they’re invisible in both framings.

If I’m serious about AI as assistive technology, I have to be serious about accessibility—not just the accessibility features AI enables, but accessibility to AI itself.

The conversation we’re not having

Here’s where I need to shift registers. Everything I’ve written so far has been framed around productivity, competition, and professional advancement. That framing is accurate—and also deeply incomplete.

Because when I think about who actually needs these tools most urgently, I don’t think about McKinsey consultants or GitHub developers. I think about my friend’s mother, who at 67 has watched her own mother and grandmother both succumb to Alzheimer’s. She doesn’t need “competitive advantage.” She needs support systems that might help her maintain independence and dignity as her memory potentially declines.

I think about my neighbor who still manages all her bills with paper files and a checkbook—not because she’s incapable of learning, but because no one has ever shown her a path from where she is to where she could be. The conversation about AI assumes a baseline of digital literacy that millions of people don’t have.

I think about the generational divide playing out in families across the country: millennials and Gen Z who grew up digital, trying to help parents and grandparents navigate a world that increasingly assumes everyone is online, everyone has apps, everyone is comfortable typing queries into chatbots.

The McKinsey framing matters for knowledge workers worried about their careers. But the most profound impacts of AI may not be in corporate productivity at all. They may be in helping a 72-year-old manage her husband’s post-stroke medical paperwork. In letting a grandfather with Parkinson’s tremors dictate messages to his grandchildren. In giving someone with early cognitive decline a system that remembers what they’re starting to forget.

Those are the stakes for millions of people. And framing AI as a corporate productivity tool completely misses this dimension.

Part IV: The Historical Pattern

Moral panic from incumbents and elites

The moral objections to AI follow a pattern so consistent across technological history that it approaches predictability.

When the printing press emerged in the 1450s, Abbot Johannes Trithemius wrote In Praise of Scribes (1494/1974), warning that printing would make monks “lazy” and “impact their souls”—the labor of hand-copying was spiritual discipline, and mechanization represented moral degradation. The delicious irony: Trithemius had his anti-printing treatise printed and was an avid collector of printed books.

When photography arrived in the 1830s, Charles Baudelaire’s famous 1859 essay in Revue Française declared it “art’s most mortal enemy,” calling photographers “failed painters, too ill-equipped or too lazy to complete their studies” (Baudelaire, 1859; see also Baudelaire, 1965).

When calculators entered classrooms in the 1970s, teachers picketed the National Council of Teachers of Mathematics annual meeting on April 4, 1986, in Washington, D.C., to protest their use, warning students would become “dependent” and lose basic computational skills (Banks, 2011). California moved in the opposite direction, prohibiting calculators on statewide assessments starting in 1997 as part of a back-to-basics movement.

John Philip Sousa coined “canned music” as an epithet for phonograph recordings, warning Congress in June 1906 that mechanical reproduction would lead to “a marked deterioration in American music and musical taste” (Sousa, 1906). He predicted, with apparent seriousness, that vocal cords would be “eliminated by a process of evolution.”

Britain’s Locomotive Acts of 1865 required a person to walk in front of automobiles waving a red warning flag and limited speeds to 4 mph in rural areas and 2 mph in urban areas (Locomotive Acts, 1861–1898). The regulations weren’t repealed until November 14, 1896, and historians credit them with setting back British automobile manufacturing by years compared to competitors (Richardson, 1977).

The consistent pattern: objections come from cultural elites, established industries, and those whose status derived from scarcity of capabilities the new technology democratized. Portrait painters objected to photography. Scribes objected to printing. Professional composers objected to recorded music. The moral arguments frequently masked economic interests, and resistance typically lasted 20–40 years before widespread adoption rendered the objections historical curiosities.

What the historical pattern doesn’t prove

I need to be honest about the limits of this argument.

Yes, calculators, photography, and printing faced moral panic that now seems absurd. But you know what else faced moral panic and turned out to be genuinely harmful? Leaded gasoline. Asbestos. Radium-painted watch dials. Thalidomide. Social media algorithms optimizing for engagement. DDT. The “they said this about X too” argument proves nothing about whether concerns about Y are valid. History contains examples of both overblown fears and prescient warnings ignored.

More specifically: the Luddites—the ur-example of irrational technology resistance—were substantially correct. Industrial machinery did devastate their livelihoods, their communities, their ways of life. Real wages for weavers fell by more than half. The transition enriched factory owners and eventually raised living standards broadly, but “eventually” meant decades of genuine immiseration for the people who lived through it. The Luddites were wrong that smashing looms would stop industrialization; they were not wrong that industrialization would harm them.

This matters because I’ve framed AI skeptics as elites protecting their status—and that framing is sometimes accurate but sometimes unfair. Some critics have substantive concerns: about job displacement, about accuracy, about concentration of power, about environmental costs, about what happens when systems trained on human data are deployed to replace human judgment. Dismissing these concerns by pointing to historical moral panics is intellectually lazy—and it’s exactly the kind of argument-by-social-position that I’ve accused skeptics of deploying.

The relevant question isn’t “did previous technologies face similar criticism?” It’s “are the specific criticisms of this specific technology valid?” That requires engaging with substance, not pattern-matching to historical analogies.

What the historical pattern does suggest is humility about our ability to predict technological impacts. The optimists and pessimists are both usually wrong about which effects materialize. The printing press didn’t destroy spirituality or social order; it did enable both the Reformation and witch-hunting manuals. Photography didn’t kill painting; it transformed what painting was for. Calculators didn’t destroy mathematical reasoning; they shifted which mathematical skills matter.

AI will likely follow this pattern: neither the utopian nor dystopian predictions will prove accurate, and the actual effects will be ones we’re not currently centered on. This argues for neither wholesale adoption nor wholesale rejection, but for the kind of thoughtful, iterative engagement that lets us learn as we go.

Who can actually afford to say no?

The contemporary AI rejection movement is concentrated in a specific demographic. A September 2024 position paper titled “Against the Uncritical Adoption of ‘AI’ Technologies in Academia” was signed by cognitive scientists and AI researchers at universities in the Netherlands, Denmark, Germany, and the US—individuals with secure positions at institutions like Radboud University, Purdue, and Cornell. Community college faculty have been explicitly told they possess “a powerful resistance tool: academic freedom”—an acknowledgment that workers without such protections cannot make the same choices.

The economic disparities are stark. IMF research notes that while AI could potentially “reduce wage inequality through the displacement of high-income workers,” these same workers are “better positioned to benefit from higher capital returns” and often find their tasks “highly complementary with AI, potentially increasing their productivity.” Companies with over $25 billion in revenue are 24–30 percentage points more likely to integrate AI than those under $100 million.

The cruel irony is that those who can afford to reject AI are precisely those whose positions are most secure, while those who cannot afford rejection—gig workers, freelancers, small business employees, people with disabilities seeking competitive employment—face pressures that make adoption survival-critical.

Part V: Taking Concerns Seriously

The environmental question: An accessibility tax

Among objections to AI, environmental concerns are often the most morally weighty. And unlike the “cheating” argument, the environmental argument is grounded in physical reality: training and running these models consumes vast amounts of energy and water.

I will not minimize this. Hannah Ritchie’s analysis estimates 2–3 grams of CO₂ per ChatGPT query, including amortized training emissions (Ritchie, 2025; see also Luccioni et al., 2024). While individual queries have a relatively small footprint—roughly equivalent to boiling water for one cup of tea—the aggregate impact is massive. As models scale, so does their resource consumption. This is a real cost.

But we need to frame this cost correctly. In medical contexts, we accept high energy costs for life-saving and life-enabling technologies. MRI machines require massive amounts of power. The production of pharmaceuticals is energy-intensive. We don’t demand that reliable medical infrastructure be carbon-neutral before we allow sick people to use it; we work to decarbonize the grid it runs on.

For disabled and neurodivergent people, AI is not a parlor trick or a toy. It is assistive technology. And like many forms of assistive technology, it comes with an energetic cost.

The question is not “is this energy use zero?” The question is “is this energy use justified by the utility it provides?”

If AI enables a neurodivergent person to maintain employment, or an elderly person to manage their independence, or a non-speaking person to communicate—that utility is high. We are paying an environmental tax for accessibility.

This doesn’t mean we give tech companies a pass. We must demand green data center energy, water-efficient cooling, and efficient model architectures. But asking individuals to reject the most powerful assistive technology of our generation on environmental grounds is asking them to pay a personal price that the broader society is not asked to pay for other high-energy luxuries (like air travel or meat consumption) that offer far less utility.

The environmental cost is real. It is the price of admission for a new layer of human capability. We should work to lower that price, but we should not deny entry to those who need it most.

The displacement question I’ve been avoiding

I’ve focused on AI as augmentation—tools that make workers more productive. But the same studies I’ve cited have an uncomfortable implication I’ve understated: if workers become 14% more productive, companies can do the same work with fewer workers.

The MIT/Stanford customer service study found that AI “weights skills toward the right tail”—meaning AI helps lower-skilled workers perform like higher-skilled workers. That’s good for lower-skilled workers who keep their jobs. But it also means companies need fewer high-skilled workers to achieve the same output. The skill premium that experienced workers command may erode.

The Harvard/BCG consultant study found that AI hurt performance on tasks outside its capability frontier. Consultants using AI on tasks AI couldn’t handle well performed worse than those working without AI—they trusted AI output that was subtly wrong. This suggests a future where workers who over-rely on AI make more mistakes, while workers who know when not to trust AI maintain an advantage. That’s a more complex picture than “AI makes everyone more productive.”

Historical patterns offer cold comfort. Previous technological transitions eventually created more jobs than they destroyed—but “eventually” often meant decades, and the new jobs weren’t necessarily accessible to those displaced. Factory work replaced artisan craft; the artisans didn’t become factory workers, they became impoverished former artisans. The gains accrued to the next generation.

I don’t have a satisfying answer here. If I’m honest, the research I’ve cited supports several interpretations:

Optimistic: AI augments human capabilities, creating new forms of valuable work while automating drudgery. Workers who learn to collaborate with AI will thrive; the transition will be broadly beneficial.

Pessimistic: AI enables companies to extract more output from fewer workers. Productivity gains will flow to capital owners, not workers. The “learn to use AI or fall behind” framing will be used to blame displaced workers for their own displacement.

Uncertain: The effects will vary dramatically by industry, role, and implementation. Some workers will benefit enormously; others will be harmed. The net effect depends on policy choices, corporate decisions, and factors we can’t yet predict.

My argument for AI adoption assumes you have choices—that you can position yourself to benefit from these tools. For workers in roles likely to be automated entirely, that advice may be cold comfort. The freelancer with ADHD might benefit from AI assistants; the call center worker whose entire job is being automated might not.

This doesn’t change my core argument that principled rejection is a luxury, but it does complicate it. For some workers, the choice isn’t “adopt AI or fall behind”—it’s “adopt AI and maybe delay displacement” versus “don’t adopt AI and definitely be displaced sooner.” Neither option is good. Telling people to embrace tools that may eliminate their livelihoods has a cruel edge I should acknowledge.

What follows? At minimum, honesty about the stakes. Advocacy for transition support, retraining programs, and safety nets for those displaced. Recognition that individual adoption choices happen within structural conditions that individuals didn’t create and can’t control. And humility about my own position—I’m writing about knowledge work from a position of relative security, and my advice may land differently for someone whose job is more directly threatened.

The accuracy problem: When AI lies convincingly

Here’s something I haven’t emphasized enough: AI systems regularly produce false information with complete confidence.

The technical term is “hallucination”—AI generating plausible-sounding content that is factually wrong, citing sources that don’t exist, or fabricating details while maintaining the tone of authoritative knowledge. This isn’t a bug being fixed; it’s an inherent feature of how these systems work. They’re trained to produce text that sounds right, not text that is right.

For the populations I’ve argued most need these tools, this is a serious problem:

The elderly parent managing healthcare: If AI confidently states an incorrect medication dosage, misidentifies a drug interaction, or fabricates appointment times, the consequences could be severe. The person who most needs help may be least equipped to catch errors.

The person with cognitive decline: External memory systems are only helpful if they’re accurate. An AI that confidently “remembers” events that didn’t happen or details that are wrong could accelerate confusion rather than providing reliable scaffolding.

The neurodivergent worker using AI for communication: If AI helps draft emails but introduces factual errors or misrepresents the user’s intentions, the tool creates problems rather than solving them.

Anyone using AI for research: The BCG consultant study found that consultants performed worse when using AI on tasks outside AI’s capabilities—specifically because they trusted AI output that was subtly wrong. Overreliance on AI can degrade rather than enhance performance.

This doesn’t negate AI’s benefits, but it fundamentally changes how we should think about adoption:

Verification isn’t optional. Every AI output needs human review, especially for consequential decisions. “AI drafted this” is the beginning of a workflow, not the end.

Critical evaluation is a skill. Using AI effectively requires knowing when to trust it, when to verify, and when to ignore it entirely. This skill takes time to develop—and the populations who most need AI assistance may have the least capacity to develop it.

Some applications are premature. Using AI to brainstorm creative ideas has different risk profiles than using AI to manage medication schedules. The assistive technology frame I’ve used implies a level of reliability that current systems don’t consistently provide.

Transparency about limitations matters. If I’m advocating AI adoption, I need to be clear that “AI can help with X” means “AI can help with X if you verify everything it produces, maintain healthy skepticism about confident-sounding claims, and never fully automate consequential decisions.”

The promise of AI as assistive technology is real. So is the risk of AI as a confidently wrong assistant that creates new problems while appearing to solve old ones. Responsible adoption requires taking both seriously.

Part VI: The Democratization Opportunity

Technology follows predictable cost curves

There’s another side to this story. AI is making capabilities accessible that were previously reserved for the wealthy and well-resourced.

Wright’s Law—the observation that costs decline predictably as cumulative production increases—is playing out dramatically in AI (Wright, 1936; Ritchie, n.d.). We’ve seen this pattern before: solar panel costs have declined 20% with each doubling of production; lithium-ion battery prices have dropped 97% since 1991. The Fortune article on AI cost collapse notes that token costs have dropped 90% in a single year, enabling startups to accomplish for $50,000 what previously required $500,000 in API credits (Maruthavanan, 2025).

Remote work platforms powered by AI-enabled tools have created 2.2 million remote workers from Brazil, Argentina, and Mexico contributing to global companies (Noceda, 2025). Software engineers in these countries earn 28% more through international remote work than local employment.

Intel reduced translation costs by 40% while doubling content volume using AI-powered localization (LILT, n.d.). A randomized controlled trial of Rori AI tutoring in Ghana found that personalized AI instruction at a marginal cost of $5 per student produced learning gains equivalent to approximately one additional year of schooling (Henkel et al., 2024)—approaching Benjamin Bloom’s famous “2 sigma” threshold for one-on-one tutoring (Bloom, 1984).

ChatGPT Plus costs $20 per month. Claude Pro costs $20 per month. These are the prices of capabilities that didn’t exist at any price point two years ago—capabilities that level playing fields for people who couldn’t afford executive coaches, writing tutors, research assistants, or coding mentors.

The question isn’t whether AI will democratize access to capability. It’s whether you’ll be among those who benefit—and whether the barriers I discussed in Part III will prevent those who most need these tools from accessing them.

On urgency and hype

I need to interrogate my own framing here.

Claims that “the competitive window is closing quickly” and “late movers risk lasting disadvantages” are the language of venture capital pitch decks and LinkedIn thought leadership. They’re designed to create fear of missing out, to pressure decisions, to suggest that deliberation is itself a form of failure. They’re also—notably—exactly the pressure tactics I’d criticize when tech evangelists use them.

The Kodak and Blockbuster analogies I might have invoked deserve skepticism. These were companies that had access to transformative technology and chose not to deploy it. That’s different from individuals making considered decisions about which tools serve their needs. And for every Kodak, there are dozens of companies that chased the next technology and failed—because the timing was wrong, the implementation was poor, or the technology didn’t deliver on its promises.

What does the evidence actually support?

Real but not universal effects: The productivity studies I’ve cited show meaningful gains for specific tasks in specific contexts. They don’t show that everyone must adopt AI immediately or face ruin.

Compound advantages exist, but so do late-mover advantages: Early adopters bear costs that late adopters avoid: buggy tools, evolving best practices, time invested in systems that become obsolete. The “optimal” adoption timing is genuinely uncertain.

Different timelines for different people: A knowledge worker whose competitors are adopting AI faces different pressures than a retiree considering whether AI might help with life management. The urgency framing collapses importantly different situations.

Hype cycles are real: Gartner’s hype cycle exists because new technologies reliably generate inflated expectations followed by disillusionment before reaching productive use. We may be at or near “peak hype” for generative AI, which would suggest that the most breathless urgency claims should be discounted.

A more honest framing: AI tools offer genuine benefits for many applications. The people best positioned to benefit are those who engage thoughtfully rather than either frantically or dismissively. There’s value in starting now because learning takes time—but there’s also value in waiting for tools to mature, best practices to emerge, and the hype to settle.

If you’re feeling pressured to adopt AI immediately or be left behind forever, that pressure is probably miscalibrated. If you’re confident that AI is a fad you can safely ignore indefinitely, that confidence is also probably miscalibrated. The truth is genuinely in the middle.

Part VII: Interdependence, Not Independence

The disability justice reframe

The disability justice framework offers a conceptual reframe that cuts through much of the moral posturing around AI tools.

Mia Mingus, a leading voice in disability justice, writes: “I am not fighting for independence... I am fighting for an interdependence that embraces need and tells the truth: no one does it on their own and the myth of independence is just that, a myth” (Mingus, 2011).

The myth of independence—that legitimate capability must be unassisted capability—reflects extraordinary privilege. As Mingus points out: someone made the clothes you’re wearing, your shoes, your car or mass transit; we don’t grow our own food. All lives are enriched, enabled, and made possible through various means of support.

This insight connects to what Angela Glover Blackwell, founder of PolicyLink, termed the “curb cut effect” (Blackwell, 2017). Accessibility features designed for disabled users benefit everyone. Voice assistants, originally accessibility features, are now universal. Nearly 90% of all video viewers use subtitles. At one Florida shopping mall, 90% of pedestrians without mobility aids went out of their way to use curb cuts rather than stairs.

When we design for disability, we make things better for everyone. When we moralize against the tools that enable participation, we hurt everyone—but we hurt the most vulnerable most.

The social model of disability, developed by the Union of the Physically Impaired Against Segregation and articulated by scholar Mike Oliver (UPIAS & Disability Alliance, 1976; Oliver, 1990, 2013), distinguishes between impairment (the physical or mental difference) and disability (the barriers society creates). AI tools don’t “fix” disabled people—they remove barriers to participation that society has constructed.

Liberatory capability vs. Corporate product

A crucial distinction is often lost in this discourse: the difference between AI as a capability and AI as a product.

When critics argue that using AI supports corporate monopolies, they are often correct about the current delivery mechanisms. OpenAI, Google, and Microsoft are indeed consolidating power. But the technology itself—the capability to run a large language model—is rapidly decoupling from these monopolies.

Open-weight models like Llama 3, Mistral, and Qwen can now be run locally on personal hardware. This is the “open source” argument applied to cognition: the ability to run these cognitive prosthetics is becoming a distributed, personal capability rather than a rented service.

For disability justice, this distinction is vital. “Interdependence” is not a metaphor here; it is a material description of how neurodivergent people use these tools. Just as a wheelchair user depends on a physical device to navigate a world built for walking, a neurodivergent user may depend on a cognitive device to navigate a world built for neurotypical executive function.

Rejecting the capability because you object to the corporate delivery mechanism is a category error. We don’t boycott wheelchairs because the manufacturing market is consolidated; we fight for better access to wheelchairs. Similarly, we can critique Big Tech’s rent-seeking while aggressively defending the liberatory power of the tools themselves.

For many disabled users, the “Intellectual purity” of rejecting AI is functionally equivalent to the “purity” of rejecting medication because “Big Pharma is evil.” The critique of the industry is valid; the rejection of the life-enabling tool is self-defeating.

We can walk and chew gum at the same time: we can adopt the tools that enable our survival and independence while advocating for open, decentralized, and ethical versions of those tools.

Part VIII: Strategic Implementation

A Mindset, Not a Manual

The transition to using AI effectively is less about learning specific software commands and more about a fundamental shift in mindset. Through observing thousands of users, I’ve seen that successful adoption consistently relies on three high-level pillars.

1. Clarity of Purpose

AI is a means, not an end. The biggest mistake new users make is asking “How can I use AI?” instead of “What specific problem am I trying to solve?”

If you can’t articulate the problem—whether it’s “I forget appointments,” “I can’t organize this essay,” or “I’m too overwhelmed to start”—you can’t effectively evaluate whether AI is helping. Start with the friction in your life, not the shiny new tool. Use AI to solve distinct, bounded problems, not as a general-purpose solution in search of a problem.

2. Rigorous Verification

The “hallucination” problem—AI’s tendency to confidently sound wrong—is not a bug you can wait for companies to fix. It is an inherent property of the technology.

Therefore, the core skill of the AI era is not “prompt engineering”; it is verification.

• The Rule: Never delegate judgment or facts to AI without a human-in-the-loop review.

• The Practice: Treat AI drafts as the work of a brilliant but unreliable intern. Helpful? Yes. Trustworthy? Absolutely not. You are the editor. You are responsible for the final output. “AI wrote this” is never an excuse for error; it is a confession of abdicated responsibility.

3. Conscious Trade-offs

Adoption involves trading one set of costs for another. You are trading privacy (data usage) for convenience. You are trading energy consumption for accessibility. You are trading the struggle of the “blank page” for the labor of editing.

These trade-offs are often worth it, but they should be made with eyes open. Be deliberate about what you are giving up to get what you need. If a use case involves highly sensitive data, the trade-off may not be worth it. If a task is critical for your own learning, the trade-off of “automating” it might rob you of necessary skill development.

Successful adoption isn’t about using AI for everything. It’s about using AI where the trade-offs lean in your favor, and ruthlessly rejecting it where they don’t.

Part IX: For Whom Rejection Might Be Reasonable

I’ve argued that rejecting AI is often a privilege. But let me be honest about cases where rejection might be entirely reasonable:

If you’ve evaluated the tools for your use case and found they don’t help. Not every application benefits from AI. If you’ve genuinely tried these tools and found they add complexity without commensurate benefit, that’s a valid reason to decline—not a moral failure.

If your field has legitimate concerns about AI use. Some professional and academic contexts have norms against AI assistance that exist for good reasons—maintaining skill development, ensuring appropriate attribution, preserving specific methodologies. Respecting those norms isn’t mere conservatism.

If you are acting in solidarity with labor. Rejecting AI tools because they are being deployed to undermine fair labor practices, replace workers without support, or devalue human craft is a rational political strategy. Protecting one’s livelihood—or standing with others protecting theirs—is a valid survival strategy, distinct from “fear of progress.”

If the privacy tradeoffs aren’t worth it for your situation. If you handle particularly sensitive information and can’t verify that AI tools meet your privacy requirements, declining to use them is prudent risk management.

If you don’t have the verification capacity. Using AI in domains where you can’t evaluate the output is genuinely risky. “I don’t know enough to catch errors” is a good reason not to use AI for that application.

If you’re prioritizing other forms of learning. A student who wants to develop writing skills through practice rather than AI assistance isn’t being foolish—they’re making a reasonable choice about skill development.

If it just doesn’t suit how you work. Individual variation is real. Some people find AI tools flow naturally into their workflow; others find them disruptive. Not every useful tool is useful for every person.

My argument isn’t that everyone must use AI tools or face moral judgment. It’s that the principled rejection of AI—the framing that treats non-use as a virtue and use as a compromise—often reflects privilege and imposes costs on those who can least afford them. Those are different claims.

The person who’s evaluated AI and chosen not to use it for specific, considered reasons is in a different position than the person who’s moralized against AI to signal sophistication while enjoying the security that makes abstention costless. I’m criticizing the latter, not the former.

Conclusion

The AI debate isn’t truly about adoption versus rejection. It’s about thoughtful implementation versus unexamined extremes—and about who gets to frame the conversation.

When tenured professors at well-funded universities sign letters against AI adoption, they’re exercising a privilege most workers don’t have. When established professionals with secure positions moralize about “authentic” work, they’re defending a status quo that benefits them. When cultural elites frame technological skepticism as intellectual sophistication, they’re repeating a pattern that stretches back centuries—and that has consistently delayed benefits for those who most needed new tools.

But I’ve also argued that adopting AI is a privilege—requiring digital literacy, hardware, connectivity, money, time, and critical evaluation skills that many people lack. The breathless enthusiasm of tech evangelists centers the experiences of the already-privileged just as much as the principled rejection of skeptics does. Both framings render invisible the people managing life on paper systems, the elders without family tech support, the workers without employer-provided tools.

The historical pattern is instructive but not dispositive. Every major information technology has faced moral panic, and that panic has typically served to delay benefits for those who most needed the technology’s democratizing effects. But history also contains technologies that faced moral panic and turned out to be genuinely harmful. The relevant question isn’t whether previous technologies faced similar criticism—it’s whether the specific concerns about AI are valid. Some are. Some aren’t. Engaging with substance requires more than pattern-matching to historical analogies.

The concerns I’ve taken seriously: Environmental costs are real, though smaller for individual use than for model training. Job displacement is real, and telling workers to embrace tools that may eliminate their livelihoods has a cruel edge I should acknowledge. Accuracy problems are real, and AI’s tendency to produce confident-sounding falsehoods creates genuine risks, especially for the vulnerable populations I’ve argued most need these tools.

Meanwhile, the people who actually need these tools—the freelancer with ADHD trying to compete with neurotypical colleagues, the researcher with chronic illness trying to maintain productivity, the aging parent trying to manage healthcare complexity, the disabled worker trying to close the employment gap—face stakes that transcend philosophical debates about technology’s soul.

The disability justice insight that all capability is interdependent—that the myth of independent, unassisted accomplishment was always fiction—offers a more honest framework. The question is not whether to use tools, but which tools, for what purposes, with what accountability, and with what safeguards against the tools’ real limitations.

For your mother with family history of Alzheimer’s, AI isn’t about staying competitive. It’s about maintaining independence, preserving memories, and managing life’s complexity even if her memory begins to fail—with clear-eyed awareness that these tools can be wrong and need verification.

For the 72-year-old overwhelmed by medical paperwork, AI isn’t about productivity metrics. It’s about reducing a three-hour weekly burden to twenty minutes—time she can spend being a caregiver instead of an administrator—while maintaining backup systems for critical information.

For the neurodivergent worker struggling with executive function, AI isn’t about gaining unfair advantage. It’s about having access to the cognitive scaffolding that neurotypical people take for granted as internal capability—with appropriate skepticism about outputs and appropriate transparency about use.

This isn’t about McKinsey or productivity—it’s about using these tools to support the very essence of being human: remembering who we are, connecting with those we love, and maintaining our dignity as we age.

The goal isn’t technological maximalism but intentional augmentation that addresses your specific needs and limitations. Whether those limitations stem from disability, resource constraints, cognitive differences, or simply the human condition of limited attention and energy—thoughtful AI implementation offers a path forward.

But that path requires honesty about risks, verification of outputs, protection of privacy, and support for those displaced. It requires acknowledging that some people’s rejection of AI is entirely reasonable for their circumstances, even as other people’s rejection reflects privilege they may not recognize. It requires taking seriously both the promise and the peril.

The question isn’t whether to adopt AI.

The question is how to implement it in ways that genuinely expand human potential—starting from wherever you actually are, addressing whatever you actually need, building toward whatever future you actually face, with clear eyes about what these tools can and cannot do.

For many, that choice was never optional.

The only question is whether we’ll approach it with the honesty, humility, and care it deserves.

References

Banks, S. A. (2011). A historical analysis of attitudes toward the use of calculators in junior high and high school math classrooms in the United States since 1975 [Master’s thesis, Cedarville University]. ERIC. https://files.eric.ed.gov/fulltext/ED525547.pdf

Baudelaire, C. (1859). Le public moderne et la photographie. Revue Française, 17, 257–265.

Baudelaire, C. (1965). The Salon of 1859: The modern public and photography. In J. Mayne (Ed. & Trans.), Art in Paris 1845–1862: Salons and other exhibitions reviewed by Charles Baudelaire (pp. 152–154). Phaidon Press.

Blackwell, A. G. (2017). The curb-cut effect. Stanford Social Innovation Review, 15(1), 28–33. https://ssir.org/articles/entry/the_curb_cut_effect

Bloom, B. S. (1984). The 2 sigma problem: The search for methods of group instruction as effective as one-to-one tutoring. Educational Researcher, 13(6), 4–16. https://doi.org/10.3102/0013189X013006004

Boston Consulting Group. (2024, October 24). AI adoption in 2024: 74% of companies struggle to achieve and scale value [Press release]. https://www.bcg.com/press/24october2024-ai-adoption-in-2024-74-of-companies-struggle-to-achieve-and-scale-value

Boston Consulting Group. (2025, September). The widening AI value gap: Build for the Future 2025. https://www.bcg.com/publications/2025/are-you-generating-value-from-ai-the-widening-gap

Brynjolfsson, E., Li, D., & Raymond, L. R. (2023). Generative AI at work (NBER Working Paper No. 31161). National Bureau of Economic Research. https://www.nber.org/papers/w31161

Brynjolfsson, E., Rock, D., & Syverson, C. (2019). The productivity J-curve: How intangibles complement general purpose technologies. MIT Initiative on the Digital Economy. https://ide.mit.edu/sites/default/files/publications/2019-04JCurvebrief.final2_.pdf

Danielson, M. L., Claussen, A. H., Bitsko, R. H., et al. (2024). ADHD prevalence among U.S. children and adolescents in 2022. Journal of Clinical Child & Adolescent Psychology, 53(3), 343–360. https://doi.org/10.1080/15374416.2024.2335625

Dell’Acqua, F., McFowland, E., III, Mollick, E. R., Lifshitz-Assaf, H., Kellogg, K. C., Rajendran, S., Krayer, L., Candelon, F., & Lakhani, K. R. (2023). Navigating the jagged technological frontier: Field experimental evidence of the effects of AI on knowledge worker productivity and quality (Harvard Business School Working Paper No. 24-013). Harvard Business School. https://papers.ssrn.com/sol3/papers.cfm?abstract_id=4573321

Doyle, N. (2020). Neurodiversity at work: A biopsychosocial model and the impact on working adults. British Medical Bulletin, 135(1), 108–125. https://doi.org/10.1093/bmb/ldaa021

Ernst & Young LLP. (2024). GenAI for accessibility: More human, not less. EY in collaboration with Microsoft. https://www.ey.com/content/dam/ey-unified-site/ey-com/en-uk/services/ai/documents/ey-gen-ai-for-accessibility-more-human-not-less.pdf

Henkel, O., Horne-Robinson, H., Kozhakhmetova, N., & Lee, A. (2024). Effective and scalable math support: Experimental evidence on the impact of an AI-math tutor in Ghana. In A. M. Olney et al. (Eds.), Artificial Intelligence in Education: AIED 2024 (CCIS, Vol. 2150). Springer. https://doi.org/10.1007/978-3-031-64315-6_34

LILT. (n.d.). Case studies. https://lilt.com/

Lloyd, G. A., & Sasson, S. J. (1978). Electronic still camera (U.S. Patent No. 4,131,919). U.S. Patent and Trademark Office. https://patents.google.com/patent/US4131919A/en

Locomotive Acts, 1861–1898 (24 & 25 Vict. c. 70; 28 & 29 Vict. c. 83; 41 & 42 Vict. c. 77).

Luccioni, A. S., et al. (2024). Power hungry processing: Watts driving the cost of AI deployment? Nature Scientific Reports. https://www.nature.com/articles/s41598-024-54271-x

Maruthavanan, T. (2025, April 4). The AI cost collapse is changing what’s possible for startups. Fortune. https://fortune.com/2025/04/04/ai-cost-collapse-tech-startups/

McElheran, K., Yang, M.-J., Kroff, Z., et al. (2025). The rise of industrial AI in America: Microfoundations of the productivity J-curve(s). MIT Initiative on the Digital Economy.

McKinsey & Company. (2024, May). The state of AI in early 2024: Gen AI adoption spikes and starts to generate value. McKinsey Global Institute. https://www.mckinsey.com/capabilities/quantumblack/our-insights/the-state-of-ai-2024

McKinsey & Company. (2025). The state of AI: How organizations are rewiring to capture value. McKinsey Global Institute. https://www.mckinsey.com/capabilities/quantumblack/our-insights/the-state-of-ai

Mingus, M. (2011, February 12). Changing the framework: Disability justice. Leaving Evidence. https://leavingevidence.wordpress.com/2011/02/12/changing-the-framework-disability-justice/

National Autistic Society. (2016). The autism employment gap. https://www.autism.org.uk/what-we-do/news/new-data-on-the-autism-employment-gap

National Safety Council. (2008). Injury facts (2008 ed.).

Noceda, P. (2025, January). E-migration is helping to fill gaps in global talent. World Economic Forum. https://www.weforum.org/stories/2025/01/immigration-e-migration-impact-nearshoring-global-talent/

Office for National Statistics. (2020). Outcomes for disabled people in the UK: 2020.

Oliver, M. (1990). The politics of disablement. Macmillan.

Oliver, M. (2013). The social model of disability: Thirty years on. Disability & Society, 28(7), 1024–1026. https://doi.org/10.1080/09687599.2013.818773

OpenAI. (2025, December 8). The state of enterprise AI: 2025 report. https://openai.com/index/the-state-of-enterprise-ai-2025-report/

Peng, S., Kalliamvakou, E., Cihon, P., & Demirer, M. (2023). The impact of AI on developer productivity: Evidence from GitHub Copilot. arXiv. https://arxiv.org/abs/2302.06590

Richardson, K. (1977). The British motor industry 1896–1939. Macmillan.

Ritchie, H. (n.d.). Learning curves: What does it mean for a technology to follow Wright’s Law? Our World in Data. https://ourworldindata.org/learning-curve

Ritchie, H. (2025, May 6). What’s the carbon footprint of using ChatGPT? By the Numbers (Substack). https://hannahritchie.substack.com/p/carbon-footprint-chatgpt

Roux, A. M., Shattuck, P. T., Rast, J. E., Rava, J. A., & Anderson, K. A. (2015). National autism indicators report: Transition into young adulthood. A.J. Drexel Autism Institute, Drexel University. https://drexel.edu/autismoutcomes/publications-and-reports/

Shaw, K. A., Williams, S., Patrick, M. E., et al. (2025). Prevalence and early identification of autism spectrum disorder among children aged 4 and 8 years—ADDM Network, 2022. MMWR Surveillance Summaries, 74(SS-2), 1–25. https://doi.org/10.15585/mmwr.ss7402a1

Social Security Administration. (n.d.). Disability and death probability tables for insured workers born in 1997 [Fact sheet]. https://www.ssa.gov/oact/STATS/disabilityTables.html

Sousa, J. P. (1906). The menace of mechanical music. Appleton’s Magazine, 8, 278–284.

Sprunt, B. (2024, July 25). A neurological disease stole Rep. Jennifer Wexton’s voice. AI helped her get it back. NPR. https://www.npr.org/2024/07/25/nx-s1-5051720/jennifer-wexton-ai-speech-progressive-supranuclear-palsy

Trithemius, J. (1974). In praise of scribes: De laude scriptorum (R. Behrendt, Trans.). Coronado Press. (Original work published 1494)

Union of the Physically Impaired Against Segregation & Disability Alliance. (1976). Fundamental principles of disability. UPIAS. https://disability-studies.leeds.ac.uk/wp-content/uploads/sites/40/library/UPIAS-fundamental-principles.pdf

U.S. Bureau of Labor Statistics. (2025, February 25). Persons with a disability: Labor force characteristics—2024 (USDL-25-0247). U.S. Department of Labor. https://www.bls.gov/news.release/disabl.htm

Wexton, J. (2024, July 24). Wexton shares video debuting new AI voice model [Press release]. Office of Congresswoman Jennifer Wexton. https://wexton.house.gov/news/documentsingle.aspx?DocumentID=952

Wright, T. P. (1936). Factors affecting the cost of airplanes. Journal of the Aeronautical Sciences, 3(4), 122–128. https://doi.org/10.2514/8.155

Yale Center for Dyslexia & Creativity. (2022). Dyslexia FAQ. Yale School of Medicine. https://dyslexia.yale.edu/dyslexia/dyslexia-faq/

There’s a moment every developer, writer, or knowledge worker eventually encounters with AI tools—the one where you suddenly realize you’ve been approaching this entire paradigm incorrectly from the beginning.

You’ve been treating Claude or GPT like a remarkably intelligent intern who unfortunately happens to suffer from complete amnesia between every single interaction. Every conversation starts from absolute zero. Every session requires the same exhausting, time‑consuming context dump where you re‑explain the fundamentals. And despite all the marketing hype about million‑token context windows and their supposed capabilities, the AI still manages to lose the thread somewhere halfway through the process of explaining your own project back to you—the very project you’ve already described multiple times.

I hit that wall hard while working on a book manuscript with a hard, non‑negotiable deadline. Thirty, then sixty‑plus chapters. Hundreds upon hundreds of research sources to synthesize and reference. A timeline that left absolutely zero room for the overwhelming cognitive overhead of constantly re‑explaining what I was even trying to accomplish, what the book’s thesis was, and where each piece fit into the larger narrative structure. Something fundamental had to change in my approach.

What I discovered through this process wasn’t a single magical tool or clever technique—it was an architecture. A completely different way of thinking about AI assistance that treats the real problem (memory limitations, context management, and coordination between systems) as the primary design constraint, not an afterthought. And 2026 is the first year where this particular architecture is not just theoretically possible, but practically achievable for everyday knowledge workers.

To understand why, you have to start with the uncomfortable truth about context.

The Real Problem Isn’t Intelligence—It’s Forgetting

Here’s what nobody really tells you when you first start working with AI agents seriously: the context window problem is far worse in practice than the technical specs suggest.

On paper, Claude can theoretically handle a million tokens in its context window. But “can handle” and “performs reliably with” are two very different things, and the gap between them matters enormously in real‑world work.

Researchers at Chroma measured 18 different large language models and found something unsettling: performance doesn’t just degrade at the edges of context windows. It becomes increasingly unreliable and unpredictable as input length grows, period. They coined a term for this: context rot.

Your AI assistant isn’t just forgetting things once you exceed some theoretical limit. It’s getting progressively less coherent, less reliable, and less accurate as the interaction accumulates context. The longer the conversation, the more fragile the model’s grip on the overall structure becomes.

Meanwhile, real work—especially in engineering, research, or serious writing—easily spans millions of tokens when you account for all the files, dependencies, documentation, and background context. Enterprise codebases, cross‑referenced research libraries, full book manuscripts plus their sources—these are orders of magnitude larger than any single context window.

The gap between what models can theoretically ingest and what real projects require is not a tiny crack you can paper over with clever prompting. It’s a canyon.

This explains why AI coding assistants, long‑form editors, or “project copilots” feel so limited once you push beyond toy problems. The issue is not that the model isn’t “smart enough.” It’s that it literally cannot hold your entire project—the architecture, the decisions, the edge cases, the style constraints—in working memory while also reasoning about what to do next.

Working memory collapses under load. Context rots. And you end up back where you started: re‑explaining the same fundamentals, over and over, just to get anything meaningful done.

So if the problem is not raw intelligence but memory and coordination, then the answer has to be architectural. You do not fix this with better prompts. You fix it by changing where memory lives.

External Tools as Extended Memory

Once you see the real constraint clearly, the solution starts to feel obvious: stop trying to cram everything into the AI’s context window. Use external tools as persistent memory systems that extend the AI’s cognitive reach.

This is what Linear, Notion, Factory, and the Model Context Protocol (MCP) actually provide when wired together correctly. Not “integrations” for a checklist. Not novelty. A genuine extension of the AI’s cognitive architecture.

• Linear becomes the task coordination layer.

It’s where work items persist reliably across sessions. Status changes survive conversation boundaries. The AI can pick up tomorrow exactly where it left off today without requiring you to reboot its entire understanding from scratch.

• Notion becomes the knowledge repository.

Specs, research documents, meeting notes, decisions, and documentation all live here. It’s too big for any context window, but it’s searchable and selectively retrievable. The AI doesn’t need all of it at once. It just needs to pull the right pieces at the right time.

• Factory’s Droids become the delegation layer.

These are autonomous agents that can work on Linear issues asynchronously while you sleep or do something else. They update their progress in Linear in a way that both humans and other agents can read and build on.

• MCP (the Model Context Protocol) is the connective tissue.

Before MCP, integrating Claude with any specific tool—Linear, Notion, your own services—required handwritten, bespoke connectors. Every new tool meant a new bespoke integration. MCP standardized this under one protocol. One interface, many tools.

Instead of shoving everything into a vulnerable, decaying context window, you build something closer to an operating system for cognition: stable storage, task tracking, standardized interfaces, and specialized workers.

Once you have that architecture, you stop fighting the model’s limitations and start exploiting its strengths.

Why Linear Specifically (and Not Just Any PM Tool)

Linear’s advantage in this stack is not just that it’s fast, though it is—benchmarks show it significantly outpacing Jira with far higher API rate limits, which matters when agents are hammering it.

The real strength is philosophical: opinionated defaults over endless flexibility.

Most project management tools brag about infinite customization. You can configure anything, define arbitrary workflows, customize fields and statuses for each team. That flexibility might sound empowering, but it quickly turns to chaos in a real organization—and it is catastrophic for AI agents.

Agents need predictable structure. Endless knobs and levers do not help; they just increase the number of patterns the agent has to infer.

Linear takes the opposite approach:

Predefined, well‑structured statuses. Clean data models. Sensible defaults that work out of the box.

When Linear notes that “flexible software lets everyone invent their own workflows, which eventually creates chaos,” they are, knowingly or not, describing why most AI agents flail in other tools. Linear is one of the rare PM systems where agents can operate with minimal custom training, because the surface area is constrained and consistent.

Then Linear went a step further: they started treating AI agents as first‑class workspace members.

A Droid or agent is not just a ghost script running in the background. It can be mentioned directly in comments, assigned issues, included on teams, and audited like any other contributor. The human remains ultimately responsible, but the AI’s work is visible, trackable, and accountable.

That design is perfectly aligned with a delegation model rather than a “magic box” model.

Factory and the Delegation Paradigm

Most AI coding tools are built around collaboration: you and the AI write code together, line by line, in something like pair programming. You stay in the loop on every micro‑decision.

That can be useful—but it is mentally exhausting if you’re trying to manage a complex project, especially in the margins of your schedule. You are still carrying the cognitive load of project management, architecture, and detailed execution.

Factory takes a different approach: delegation.

You define a complete task with clear objectives. The AI (a “Droid”) executes autonomously. You review the result.

This model shifts your role upward:

• You define problems, constraints, and priorities.

• The Droid handles the mechanics of execution.

• You return for review, not for every intermediate step.

Factory’s branding choice—calling them “Droids” instead of “agents”—is not just cute. It’s a deliberate distancing from the hype‑laden, brittle behavior of typical “agents” that spiral into while‑loops. A droid is meant to be a specialized worker you can rely on.

Their benchmark results are not just marketing slides. Factory’s Droids hit top position on Terminal‑Bench, a standard for evaluating coding agents, beating well‑known baselines by a meaningful margin. One case study described a hospital system migration that went from four months with ten people to three and a half days with two engineers plus Droids.

The underlying technique they call “four‑layer context engineering” treats context the way a good OS treats memory: finite, managed, budgeted. Not something you just max out until it breaks.

In other words, Factory internalizes the same lesson: memory and context must be engineered, not assumed.

A Workflow That Actually Changes How Work Feels

A solo developer named Dylan Goings wrote about the emotional side of this architecture in a way that hit me hard.

He wasn’t drowning because his side projects were technically impossible. He was drowning in cognitive overhead:

• Constantly switching roles and mental models.

• Rebuilding context every time he sat down for an hour.

• Losing momentum because the cost of resuming work was so high.

His solution was to shove all that overhead into an external system:

• He writes high‑level stories in Linear describing what needs to be built.

• An AI agent refines them into detailed implementation plans.

• Remote agents work asynchronously on those tasks, updating statuses and fields as they go.

• Status changes act as coordination locks to prevent agents from overlapping or colliding.

The result is not just efficiency. It was a psychological shift. He described the experience as re‑discovering the joy of building software—because his limited time could finally go toward design and creativity, not re‑orientation.

This is the deeper point: once your architecture solves memory and coordination, it doesn’t just change what you can do. It changes how it feels to work at all.

All Actions in Linear, All Notes in Notion

The cleanest practical rule of thumb for this stack comes from Climate Policy Radar’s real‑world practice:

All actions in Linear. All notes in Notion.

This is not arbitrary. It reflects a division of labor that aligns both with human needs and AI constraints.

• Linear handles execution.

Sprints, issue tracking, status, GitHub integration. It’s the operational heartbeat of getting things done.

• Notion handles knowledge.

Specs, research, meeting notes, long‑form thinking. All the information you need to reference, but don’t want cluttering your immediate execution environment.

The integration between them is simple but powerful: paste a Linear URL into Notion, and you get a live preview of the issue status and fields. On higher‑tier plans, Notion’s AI can query Linear directly. For more advanced setups, tools like Unito can keep the two in tighter sync.

You are not asking one tool to be everything. You are letting each do what it’s best at, and letting MCP and AI bridge the gap.

That leads naturally to the question: where does your own local tooling fit in? What about Obsidian, git, and the comfort of local‑first workflows?

From Local‑First to Cloud‑Scale

This is where my own morning notes came into focus: I’ve been wrestling with a tension between local‑first tools like Obsidian and the raw computational power available in the cloud.

For a long time, my personal knowledge management revolved around Obsidian, backed by a git workflow on my desktop. I manage final drafts in markdown, open them with multiple tools, and use version control to track iterations. It’s fast, transparent, and robust. I know exactly where my files live.

But there’s a hard limit to what a single machine can do.

Unless you own an extremely high‑end rig, the parallel processing you get from cloud compute utterly dwarfs what a personal PC can deliver—even a $10,000 custom build. You can throw as much money as you want at local hardware, but you will not match the elasticity and parallelism of fleets of GPUs working in the background.

So I found myself at a crossroads:

• Obsidian is excellent. I’m not uninstalling it.

• My git‑backed markdown workflow is still deeply satisfying.

• But the center of gravity is shifting toward where my data can be most effectively processed and queried, not just where it sits.

And that is what pushed me toward a different way of thinking about “local‑first.”

Why Obsidian’s Portability Still Matters

Here’s the key realization: I did not adopt Obsidian because it was the perfect app. I adopted it because everything in it is fundamentally portable.

Markdown plus git is not a brand. It is a guarantee.

• I can migrate those notes elsewhere.

• I can re‑organize them into a wiki.

• I can feed them to another system entirely.

• I can bring them back if I ever need to.

The format is the asset. The platform is contingent.

Reliable persistent data is all that matters. As long as I can extract and transform it, the choice of editor becomes tactical, not existential.

So my goal shifted: use Obsidian as a staging ground, but design agents to migrate and re‑structure that knowledge into coherent, queryable wikis in Notion, where it can plug directly into the cloud compute and AI stack I’ve been building.

Notion becomes the surface where:

• My knowledge is indexed and cross‑linked.

• My AI tools (via MCP) can instantly search across it with minimal latency.

• That same knowledge is reachable by agents orchestrated through Linear or specialized tools like Factory.

If I can point an agent at Notion instead of at loose folders of markdown, I inherit all the indexing and connective tissue automatically. And because the source material remains portable markdown, I haven’t sacrificed long‑term flexibility.

The conclusion is subtle but important: Obsidian’s value is not diminished. It has just moved “down a layer” in the stack. It is my local canvas and raw material store. Notion and the cloud stack are where that material is compiled into living systems.

Cloud Compute vs. Local Hardware

This leads naturally to a blunt economic comparison.

I can pour thousands of dollars into a single local machine to get better CPU and GPU performance. Or I can spend a fraction of that on a Claude Max subscription, or equivalent frontier model access, and tap into infrastructure that vastly outperforms what I could ever reasonably own.

When you factor in:

• massive parallelism for multi‑file tasks,

• specialized inference hardware, and the

• ongoing improvements in model quality,

it becomes clear that trying to compete with cloud compute at home is a losing game, except for very specific edge cases.

So the question becomes: if I have portable data (markdown, git), and I can move it into tools like Notion that are wired to cloud‑scale AI through MCP, why would I chain my workflows to the limits of my personal hardware?

I still want the safety, transparency, and control of local files. But I want the brains and horsepower of the cloud.

For a while, that trade‑off felt like a permanent compromise.

And Then Cowork Dropped Today

Cowork is what finally snapped this architecture into a new configuration.

Up to this point, there was always a seam between local and cloud workflows. I could:

• Keep my files in Obsidian, backed by git.

• Use Notion as my cloud‑native knowledge surface.

• Use Linear and Factory as my orchestration and delegation layers.

• Use Claude, GPT, and others as the engines.

But there was friction every time I needed the AI to directly manipulate my local files. I still had to:

• Manually run scripts or commands.

• Shuffle files between environments.

• Babysit some steps of the pipeline because the model itself did not have direct, autonomous access to my filesystem.

Cowork changes that.

Cowork gives Claude direct, persistent, permissioned access to specific folders on my machine. It can:

• Read, edit, and create files.

• Manage versions in a way that cooperates with git.

• Orchestrate multi‑file transformations autonomously while I do something else.

This doesn’t just smooth out my git workflow. It automates the exact migration pipeline I had been planning manually:

• Take sprawling Obsidian vaults.

• Read all the markdown.

• Normalize, restructure, and tag content.

• Generate structured outputs—Notion‑ready pages, wikis, or content bundles.

• Do it repeatedly and systematically, without me clicking through every file.

I can queue these tasks as if I’m leaving instructions for a coworker:

• “Clean up and categorize this folder.”

• “Convert these notes into a structured wiki outline.”

• “Cross‑link related files and summarize the relationships.”

• “Prepare this subset of notes as a Notion import, following this schema.”

While Cowork operates, I am free to focus on higher‑level thinking, instead of dragging files around and copy‑pasting between tools.

And here’s the deeper consequence: this setup validates the entire cloud‑over‑hardware argument. I now have:

• Enterprise‑grade AI running in the cloud.

• Permissioned, programmatic access to local files.

• A portable knowledge format (markdown and git).

• Notion as the cloud‑native wiki and knowledge surface.

• Linear and Factory handling orchestration and delegation.

• MCP stitching the whole thing together under a single protocol.

I am no longer forced to choose between “local safety + control” and “cloud power + integration.”

Cowork becomes the bridge.

One Cohesive Architecture

Put together, this is the workflow that actually works in 2026:

Local‑First Creation and Raw Notes

Drafts, scratch work, and ad‑hoc notes live in Obsidian as markdown.

Git manages version history.

Files remain fully portable and tool‑agnostic.

Cloud‑Scale Processing with Direct Local Access

Cowork gives Claude direct access to selected folders.

Claude can batch‑process, reorganize, and transform entire directories.

Multi‑file refactors, migrations, and reorganizations become trivial and parallelizable.

Structured Knowledge in Notion

The most valuable, reusable knowledge is migrated into Notion.

Pages become canonical specs, wikis, research hubs, and documentation.

Notion MCP connects this knowledge base directly to all my AI providers.

Orchestrated Work in Linear

Every meaningful action becomes a Linear issue or subtask.

AI creates and refines task breakdowns.

Human and agents share the same source of truth about “what’s next.”

Delegated Execution via Factory and Other Agents

Droids handle encapsulated work items.

Status updates, logs, and results land back in Linear and the relevant repos.

I review and steer, instead of micromanaging.

MCP as the Universal Connector

Rather than building one‑off integrations, everything talks through MCP.

New tools plug into the same protocol.

The cognitive architecture stays stable even as individual tools evolve.

Zero Manual PM Work as the End State

I can feed a chaotic draft or half‑formed project into Claude.

It can explode that into a detailed roadmap inside Linear.

Agents and Droids take it from there, updating everything as they go.

Notion, Linear, and my repos stay in sync through a mix of native integrations, MCP, and targeted automation.

In this world, the AI doesn’t “remember” my project because it has great memory. It remembers because I’ve externalized memory into systems designed for persistence, coordination, and search.

The stack is not magical. It is intentional:

Linear for orchestration.

Notion for knowledge.

Obsidian and git for raw, local, portable creation.

Cowork as the bridge between local files and cloud intelligence.

Factory and similar tools for true delegation.

MCP as the universal connector tying it all together.

Your particular tools might differ. Maybe you use different repos, different agents, or different cloud models. But the underlying principles are stable:

Keep your data portable.

Put long‑term memory in external systems, not in a context window.

Separate knowledge from action.

Use opinionated, structured tools where agents need to operate.

Delegate execution. Retain direction.

The Killer Feature: Big‑Picture Work, Not Micro‑Management

The most important shift in this architecture is not technical. It’s cognitive.

Once you accept that:

• project management drudgery can be entirely automated,

• file migrations can be handed to agents,

• multi‑tool coordination can flow through MCP and well‑designed APIs, and

• local vs. cloud is no longer an either/or decision,

you suddenly regain the mental space to do big‑picture work.

Coding stops being the marquee skill. Configuration and integration stop being the bottlenecks. The scarce resource becomes:

• architectural thinking,

• system design,

• clear problem statements,

• and the willingness to keep simplifying the path from “stuck project” to “finished thing.”

The workflow that actually works in 2026 is not about finding “the perfect AI tool.” It is about assembling a small, coherent stack where:

• All knowledge lives in a form that is search‑friendly and portable.

• All actions are coordinated through a single, opinionated system.

• AI handles everything in between.

So if you take only one concrete action from this entire narrative, let it be this:

• Start treating your tools as parts of an external brain instead of as single, magical apps.

• Put all actions in a structured system like Linear.

• Put your durable knowledge in Notion (or an equivalent that your agents can query easily).

• Keep your raw notes in portable formats like markdown.

• Let agents, Droids, and Cowork do the mechanical work of moving information between those layers.

• All knowledge in Notion.

• All actions in Linear.

• Your raw, local brain in Obsidian.

• Cloud agents and Cowork handling everything in between.

That is the workflow that actually works in 2026.

Selected Reading

Chroma on “context rot” – Long context, short memory: a study of LLM context behavior

https://www.trychroma.com/blog/long-context-short-memory

Anthropic – What is the Model Context Protocol (MCP)?

https://www.anthropic.com/news/model-context-protocol

MCP on GitHub

https://github.com/modelcontextprotocol

Linear – product overview

https://linear.app

Factory – Droids and AI software factory

https://factory.ai

Anthropic – Cowork overview

https://www.anthropic.com/news/cowork

Climate Policy Radar – All actions in Linear, all notes in Notion (Linear + Notion setup)

https://climatepolicyradar.org/blog/linear-notion-setup

Dylan Goings – The workflow that made side projects fun again

https://dylangoings.com/posts/ai-workflow

What began as a brief interruption in my normal routines a couple months ago has progressed into a necessary hibernation from posting actively here on Substack.

The short version of the story is that healing from my recent bottom surgery revision is taking its time. It has taken a significant toll off my physical, mental, and financial health. It’s time to call this break what it is.

I look forward to sharing more with you all in the new year when I’m in a better space to maintain a regular posting schedule. For now, I’m recognizing the wisdom within myself calling for restraint and taking some things off my plate. I have paused all billing for paid subscribers. I want to say more, but this will have to do for now.

Be kind to yourselves while I’m away. ✨

Many blessings,

Axl

Introduction

For many transgender, non-binary, and altersex people, the journey toward living an authentic life involves aligning their physical body with their internal sense of self. Vaginoplasty is a profound and often essential step in this process; it is a life-affirming experience designed to construct a neovagina and often the surrounding external anatomy, known as the vulva.^1,3

This process is a deeply personal one, and it is important to recognize that there is no single “right” way to transition or to affirm one’s gender.⁵ The path is as unique as the individual traveling it, and vaginoplasty represents one of many options available to help create a body that feels like home.

The Goals of Surgery: Beyond the Physical

The decision to pursue vaginoplasty is driven by a range of personal goals that extend far beyond the purely physical. The primary purpose of the surgery is to alleviate gender dysphoria—the significant distress that can occur when one’s physical characteristics do not match their gender identity.³ By creating anatomy that is congruent with one’s identity, the surgery can lead to a profound sense of relief, wholeness, and peace.

Beyond this central aim, the intended results of vaginoplasty are multifaceted. A key objective is to create an aesthetic appearance that feels natural and authentic to the individual, fostering a positive body image and self-confidence.³ Great care is taken to preserve and repurpose sensitive nerve tissues to create a neoclitoris, thereby maintaining the capacity for sexual sensation and orgasm. For those who desire it, the procedure aims to create a vaginal canal of sufficient depth and width to enable comfortable and pleasurable receptive penetrative sex.³ Furthermore, for individuals who have not already had an orchiectomy (removal of the testes), this procedure is a standard component of nearly all vaginoplasty techniques. This removal eliminates the body’s main source of testosterone production and can reduce or end the need for androgen-blocking medications for transfeminine individuals.¹

The evolution of medical understanding has shifted the focus of this surgery from a simple anatomical reconstruction to a holistic, identity-affirming procedure. This change is reflected in the very language used, moving from terms like “sex reassignment” to “gender affirmation”.⁶ This modern approach recognizes that the success of the surgery is measured not just by the technical skill of the surgeon, but by its transformative impact on a person’s mental, emotional, and overall quality of life. It is a process designed to support an individual’s well-being in the most fundamental way.

A Note on Terminology: Respectful and Accurate Language

Throughout this guide, an effort will be made to use anatomical terms that are both respectful and medically accurate.⁵ The newly created vagina is called a neovagina. The external genitalia, including the labia, clitoris, and urethral opening, are collectively known as the vulva. The term vaginoplasty refers to procedures that create a neovagina. A vulvoplasty is a procedure that creates only the external vulva, without a vaginal canal. A labiaplasty is a procedure to create or reshape the labia (the inner and outer “lips”).⁵ While these are the standard terms, it is important to remember that every individual has the right to use the language that feels most comfortable and affirming for their own body.

Subscribe now

Exploring the Spectrum of Surgical Techniques

The Philosophy of Choice: There is No “One-Size-Fits-All”

The field of gender-affirming surgery has evolved significantly, and today there is a spectrum of surgical techniques available to meet a wide variety of individual goals, anatomical considerations, and health profiles.⁵ This diversification of options represents a fundamental shift in transgender healthcare, moving away from a single, binary model of transition and toward a truly patient-centered approach. The choice of technique is a deeply personal decision made in partnership with a surgeon, based on a holistic view of a person’s desires and needs.² The existence of varied procedures empowers individuals to design a body that aligns with their unique gender identity, whether that falls within or outside of the gender binary.

The Foundational Technique: Penile Inversion Vaginoplasty (PIV)

Penile inversion vaginoplasty is the most common and widely performed technique for creating a neovagina.² It is a highly refined procedure that skillfully repurposes existing genital tissues. The process begins with the surgeon carefully creating a new space, or pocket, for the vaginal canal. This space is located between the rectum behind and the bladder and prostate in front, nestled between the pelvic floor muscles.³

The lining of this new canal is created primarily from the skin of the penile shaft, which is inverted (turned inside out) like a sock.³ If additional depth is needed, or if an individual was circumcised and has less penile skin available, surgeons use a skin graft—often from the scrotum—to extend the lining. Scrotal skin is also used to construct the labia majora, the outer lips of the vulva.³ The scrotal skin is also used to construct the labia majora, the outer lips of the vulva.¹¹

One of the most critical steps in the procedure is the creation of the neoclitoris. A small, sensitive portion of the glans (the head of the penis) is meticulously preserved with its original nerve and blood supply and repositioned to function as a clitoris, with the goal of retaining erotic sensation.² The urethra is shortened and its opening is moved to a new position within the vulva.⁴ The prostate gland is not removed; leaving it in place helps prevent urinary complications like incontinence and can also be a source of pleasurable sensation for some individuals.⁴ The success of this technique relies on having a sufficient amount of healthy, elastic genital tissue, which can sometimes be a consideration for those who started hormone therapy at a very young age and may have experienced less genital development.²

Innovations for Self-Lubrication and Depth

While penile inversion is the standard, surgeons have developed innovative techniques that use different tissues to address specific goals, such as achieving greater vaginal depth or creating a self-lubricating canal.

Peritoneal Pull-Through (PPT) Vaginoplasty: This advanced technique utilizes the peritoneum, which is the thin, moist membrane that lines the abdominal cavity.² During the procedure, a portion of this lining is gently pulled down to form the upper part of the neovagina. The peritoneal tissue offers several distinct advantages: it is naturally pink in color, it is self-lubricating, secreting a clear fluid similar to natural vaginal lubrication, and it does not grow hair.⁹ This can reduce the need for external lubricants during sexual activity and may result in a neovagina that feels and functions more like a cisgender vagina. However, it is important to understand that PPT is a more complex surgery. Because it involves entering the abdominal cavity, it carries additional risks compared to a standard penile inversion and requires a surgeon with specialized training in this technique.⁹

Colovaginoplasty (Intestinal Vaginoplasty): For individuals who require significant vaginal depth, or in cases where there is not enough local tissue for a penile inversion (such as in some revision surgeries), surgeons may perform a colovaginoplasty.² This procedure uses a segment of the large intestine, most commonly the sigmoid colon, to construct the neovagina. Like the peritoneum, intestinal tissue is naturally moist and self-lubricating, and it can be used to create a vaginal canal of considerable and stable depth.² This makes it an excellent option for achieving specific functional goals. The trade-offs include a more extensive surgery with a longer recovery period, the risks associated with any bowel surgery, and the potential for excess mucus production that may require management.²

The Vulvoplasty Option: Aesthetics Without a Vaginal Canal

Gender affirmation is not always about creating function for penetrative sex. For many, the primary goal is to have external genitalia that look and feel right. Vulvoplasty, sometimes called “zero-depth” or “shallow-depth” vaginoplasty, is a surgical option that masterfully addresses this need.² This procedure involves all the aesthetic steps of a full vaginoplasty—creating a clitoris, labia majora, and labia minora—but without creating an internal vaginal canal.⁴ The outward appearance is nearly identical to that of a full-depth vaginoplasty.

The most significant advantage of choosing vulvoplasty is that it completely eliminates the need for vaginal dilation.² Dilation is a demanding, lifelong commitment required to maintain the depth and width of a neovagina. By forgoing the creation of a canal, individuals who choose vulvoplasty can have a significantly simpler and often faster recovery. This makes it an ideal choice for those who have no interest in receptive vaginal intercourse, or for anyone who wishes to avoid the rigorous post-operative dilation regimen for personal or medical reasons.² This choice fundamentally decouples the aesthetic affirmation of having a vulva from the functional requirements of maintaining a vaginal canal, representing a powerful affirmation of individual autonomy in the transition process.

Embracing the Power of “And”: Phallus-Preserving Vaginoplasty (PPV)

The evolution of gender-affirming surgery is perhaps most powerfully illustrated by the development of phallus-preserving vaginoplasty (PPV), also known as penile-preserving vaginoplasty or, in some communities, “Salmacian surgery”.⁸ This procedure challenges the binary assumption that one must choose between having a penis or a vagina, offering a vital and affirming option for many non-binary, genderqueer, and other gender-diverse individuals who desire to embrace the power of ‘and’ by having both.⁹

In a PPV, a neovagina is created while the existing phallus is left intact.⁸ Because the penile skin is not used to line the new vaginal canal, the surgeon must use other tissues. This is often accomplished using the peritoneal pull-through (PPT) technique described earlier, which provides a self-lubricating lining.⁹ Alternatively, the canal can be created using scrotal tissue or skin grafts taken from other areas of the body, such as the lower abdomen, hip, or inner thigh.⁷

The procedure can be customized to a great degree based on the individual’s goals. PPV allows individuals to maintain the sexual function, appearance, or personal sense of identity associated with their penis while also having a functional vagina, offering a way to align their body with their gender identity in a manner that feels uniquely right for them.⁸

Overview of Vaginoplasty Techniques

The Path to Surgery: Candidacy, Consultations, and Personal Goals

Determining Your Candidacy

The journey to vaginoplasty begins with determining candidacy, a process that involves both personal readiness and medical evaluation. A good candidate is someone who identifies as transgender or non-binary, experiences gender dysphoria, and has a persistent and well-documented desire to align their physical body with their gender identity, while also being in good overall physical and mental health.¹ It is also essential to be in good overall physical and mental health. This is not a barrier to care, but rather a crucial safety measure to ensure the body is prepared to undergo a major surgical procedure and to minimize the risk of complications during and after the operation.¹

Understanding the WPATH Standards of Care

Most surgeons and insurance providers adhere to the guidelines established by the World Professional Association for Transgender Health (WPATH).¹⁶ These Standards of Care are designed to ensure that individuals are well-prepared and supported throughout their transition. A key component of these guidelines is the requirement for letters of support from qualified mental health professionals.⁶

These letters are not meant to “gatekeep” access to care. Rather, they serve two important purposes. First, they document an individual’s gender dysphoria and their capacity to provide informed consent for the surgery. Second, and more importantly, they confirm that the person has explored the implications of surgery and has the psychological readiness to navigate the significant physical and emotional changes that lie ahead.⁶ The process of working with a therapist to obtain these letters is an opportunity to build a strong therapeutic relationship that will be an invaluable source of support during recovery. Insurance companies may have additional requirements, such as a documented period of living in a gender role congruent with one’s identity or a certain duration of hormone therapy.⁵

The Consultation: A Partnership with Your Surgeon

Selecting a surgeon is one of the most important decisions in this journey. It is advisable to seek referrals from trusted sources, such as support groups or other transgender individuals, and to thoroughly research a surgeon’s credentials, experience, and results.¹ The surgical consultation should be viewed as the beginning of a partnership. It is a collaborative dialogue where the individual can openly express their hopes, fears, and goals.¹

A good surgeon will listen carefully, conduct a thorough medical history and physical examination, and provide a clear and honest assessment of what can be achieved.⁵ They will explain the different surgical options, the potential risks and benefits of each, and help the individual understand what to expect in terms of aesthetic and functional outcomes.¹ This is the time to ask questions, voice concerns, and ensure a feeling of comfort and confidence with the surgeon and their team. A strong, trusting relationship with the surgical team is a cornerstone of a positive experience.

Defining Your Personal Surgical Goals

The pre-operative journey is not simply about meeting external requirements; it is a profound process of self-reflection and decision-making. Before and during consultations, it is helpful to think deeply about one’s personal priorities for the surgery.¹⁷ This process of defining goals is essential because it helps tailor the surgical plan to the individual, ensuring the chosen procedure is the best fit for their unique vision of their body and life.

Common priorities to consider include ⁵:

• Aesthetics: What is the desired appearance of the vulva? Is symmetry important? What are the preferences for the size and shape of the labia?

• Sexual Function: Is the ability to have receptive penetrative sex a primary goal? How important is achieving orgasm? What are the expectations for sexual sensation?

• Urinary Function: Is there a preference for how and where the urinary stream will be directed?

• Recovery and Long-Term Maintenance: What is the willingness and capacity to engage in the lifelong dilation regimen? Is minimizing recovery time a high priority?

By clarifying these goals, an individual can work with their surgeon to select the most appropriate technique—whether it be a full-depth penile inversion for penetrative function, a vulvoplasty to prioritize aesthetics and avoid dilation, or a phallus-preserving procedure to honor a non-binary identity. This entire pre-operative phase, from therapy to consultations to self-reflection, functions as an integrated system. It is designed not as a series of hurdles, but as a comprehensive process of preparation and risk mitigation, ensuring that by the time surgery arrives, the individual is physically, emotionally, and logistically prepared for the best possible outcome.

Preparing for a Successful Transformation

Preparing Your Body and Mind

The period leading up to surgery is not a passive waiting time; it is an active and essential phase of the treatment itself. The commitment and actions taken during these weeks and months are just as critical to a safe and successful outcome as the surgeon’s skill in the operating room. This preparation is a holistic process, involving the optimization of physical health, the completion of necessary medical prerequisites, and the fortification of one’s emotional and psychological resilience.

Physical Health Optimization: Building a Strong Foundation

Preparing the body for the stress of major surgery is one of the most powerful steps an individual can take to promote smooth healing and reduce the risk of complications. This involves adopting and maintaining healthy lifestyle habits, including eating a nutritious, well-balanced diet and engaging in regular physical exercise to improve overall fitness and cardiovascular health.¹

Of paramount importance is the strict cessation of all forms of smoking and nicotine consumption, as nicotine is a potent vasoconstrictor that narrows blood vessels and severely restricts blood flow to tissues.³ This impairment of circulation can dramatically delay wound healing and lead to serious complications, such as tissue death (necrosis).³ For this reason, surgeons require patients to be completely nicotine-free for at least three months prior to surgery and to remain so for at least six weeks afterward.³ Nicotine tests are often performed to confirm compliance.¹⁵

Similarly, it is crucial to avoid the use of illicit substances such as cocaine, methamphetamines, and heroin, as they can negatively affect healing and pose extreme dangers when interacting with anesthesia.⁵ Being completely honest with the surgical and anesthesia teams about all substance use, as well as alcohol consumption, is not a matter of judgment but of absolute safety.⁵

The Critical Step of Hair Removal

For most vaginoplasty techniques that create a vaginal canal, permanent hair removal from the genital area is a mandatory prerequisite.¹¹ This is because tissues that naturally grow hair, such as the skin of the scrotum and the base of the penis, are often used to construct the neovagina. If hair follicles are not permanently destroyed before surgery, hair can continue to grow inside the vaginal canal.³ This internal hair growth is not only a source of potential irritation, discomfort, and infection but can also be a significant source of ongoing gender dysphoria for many individuals.⁵

It is important to note, however, that not all vaginoplasty techniques require extensive permanent hair removal. For example, some techniques, such as Peritoneal Pull-Through (PPT) vaginoplasty, utilize peritoneal tissue from the abdominal cavity to line the neovagina. This tissue does not grow hair, thus often eliminating the need for pre-surgical hair removal in the genital area. It is crucial to discuss the specific requirements of your chosen surgical technique with your surgeon during your consultation.

The two primary methods for permanent hair removal are laser hair removal and electrolysis. The surgeon will specify which areas require treatment, but this typically includes the entire scrotum, the shaft and base of the penis, and the perineum (the area between the scrotum and anus).¹¹ It is vital to understand that this process takes a significant amount of time. Achieving complete and permanent hair removal can require multiple sessions spread out over 12 to 18 months.⁵ Because of this long timeline, it is strongly recommended to begin the hair removal process as early as possible, often long before even having the initial surgical consultation.⁵

Adjusting Your Hormone Regimen

As surgery approaches, your medical team will provide specific, individualized instructions regarding your medications, including any hormone therapy. Protocols vary between surgeons and are tailored to your specific health profile, hormone regimen, and surgical goals, so it is essential to follow your team’s guidance precisely.

For individuals on feminizing hormones, a common practice is to temporarily stop taking estrogen and any progesterone-containing medications to reduce the risk of developing blood clots (venous thromboembolism), a potential complication of major surgery.³ This pause usually begins about two weeks before the surgery date.³ However, this is not a universal requirement, and some surgeons’ protocols may allow for continuation of hormones, so it is vital to clarify this with your specific surgical team.

For individuals on masculinizing hormones, such as testosterone, the protocol may be different. Testosterone therapy is often continued throughout the surgical period. Your surgical team will manage any associated risks, such as polycythemia (a high concentration of red blood cells) or an increased risk of blood clots, with measures like VTE prophylaxis. It is crucial to bring all your medications, including testosterone, with you to the hospital and discuss your regimen with the medical team.

It is also important to note that while many surgical pathways include at least 12 months of hormone therapy, it is not always a strict requirement for surgery. Some individuals may be eligible for surgery without having been on hormones, especially if hormone treatment is not desired or is medically contraindicated. For those undergoing vaginoplasty without prior or current testosterone suppression, it is important to discuss the potential for hair to grow inside the neovagina if hair-bearing skin is used in its construction.

Since all vaginoplasty techniques include an orchiectomy for patients who have not had one previously, your body’s natural hormone production will be significantly altered. After surgery, your hormone regimen will need to be re-evaluated and adjusted by your endocrinologist or hormone provider to find the appropriate maintenance dose that aligns with your long-term health and transition goals, which may include estrogen, testosterone, or other options.

Emotional and Psychological Preparation: Your Inner Toolkit

Just as the body must be prepared for surgery, so too must the mind and spirit. The emotional journey of transition is complex, and surgery is a major milestone that can bring a whirlwind of feelings. Mental health is a vital component of a successful surgical experience.¹⁴

Ongoing psychological support from a therapist who is knowledgeable about transgender health is strongly recommended, both in the months leading up to surgery and throughout the recovery period.⁶ Therapy provides a safe and confidential space to process anxieties, manage expectations, and develop coping strategies for the challenges of recovery.¹⁹

Building a robust support system is equally crucial. This network can include affirming friends, chosen or biological family, and community support groups.¹ Connecting with peers who have been through similar experiences can be incredibly validating, offering practical advice and a powerful reminder that one is not alone on this path.¹⁹

Another valuable preparatory step is engaging with a pelvic floor physical therapist. These specialists can provide education about the pelvic anatomy, teach exercises to help relax and control the pelvic floor muscles, and provide strategies for managing bladder control and optimizing sexual function after surgery. Beginning this work before the operation can lead to a smoother, more confident recovery.⁵

Logistics and Practical Planning

A successful recovery from vaginoplasty is not just a medical event; it is an ecosystem that depends on a well-planned environment of social, physical, and financial support. The logistical preparations made before surgery are a critical investment in a safe and comfortable healing process. The understanding that a patient’s recovery environment directly impacts their outcome has led medical teams to view this planning as an essential part of the surgical journey itself.

Arranging Your Support System

For at least the first four to six weeks after surgery, an individual’s mobility, energy, and ability to perform daily tasks will be severely limited; therefore, having a reliable and compassionate support person—or a team of helpers—is not a luxury, but a necessity.¹⁵ This support system is vital for both physical assistance and emotional well-being.

Practical tasks that a caregiver can assist with include ¹⁸:

• Transportation: Providing rides to and from the hospital and to all follow-up appointments.

• Household Management: Taking care of grocery shopping, preparing nutritious meals, washing dishes, doing laundry, and managing other household chores.

• Personal Care: Assisting with errands like picking up prescriptions and, in the very early days, potentially helping with basic hygiene or wound care as directed by the surgical team.

• Dependent Care: Looking after any children, pets, or other family members who rely on the individual for care.

• Companionship: Simply being present to offer emotional support, provide encouragement, and help combat the feelings of isolation that can arise during a long recovery period.

Preparing Your Home for a Safe Recovery

Modifying the home environment before leaving for the hospital can make the return home significantly easier and safer. The goal is to create a “recovery nest” that is comfortable, accessible, and minimizes the risk of falls or strain.²²

Helpful home modifications include ²²:

• Create Clear Pathways: Remove any clutter, rugs, or electrical cords that could become tripping hazards.

• Establish a Recovery Hub: Set up a comfortable space for rest, preferably on the main floor to avoid stairs. A cart or bedside table should be stocked with all essentials to keep them within easy reach. This includes medications, a large water bottle, healthy snacks, phone and chargers, books, hygiene products, extra towels, and the remote control.

• Adapt the Bathroom: A shower chair can be invaluable, as standing for long periods may be difficult. Temporary grab bars can provide stability and prevent slips.

• Prepare for Rest: Wash all bedding and have clean, loose-fitting, and comfortable clothing easily accessible.

• Stock the Kitchen: Prepare and freeze meals in advance. Stock up on non-perishable groceries, canned goods, and easy-to-prepare foods. Consider setting up a grocery or meal-kit delivery service for the recovery period.

Financial and Work-Related Planning

Vaginoplasty requires a significant period of time away from work and daily responsibilities. Planning for this interruption is a key part of reducing stress during recovery. Most individuals should anticipate needing approximately six to eight weeks off from work to allow for proper healing and to regain strength, so it is important to formally request this time off from an employer as far in advance as you can.³

A thorough financial plan should also be put in place.¹⁸ This involves budgeting for regular living expenses that will continue during the recovery period, such as rent or mortgage payments, utilities, and food. It is also essential to account for additional surgery-related costs, which may include ¹⁸:

• Medications and medical supplies not covered by insurance.

• Travel and accommodation expenses, especially if the surgery is out-of-state.

• Potential costs incurred by the support person, such as their own travel or time off from work.

Packing Your Hospital Bag and Assembling Home Supplies

Being organized with supplies for both the hospital stay and the return home can provide a sense of control and preparedness.

For the hospital bag, consider packing³:

• Loose, comfortable clothing to wear home. An easy-to-slip-on dress or skirt is often ideal for discharge day.

• Non-skid slippers with a rubber sole for walking in the hospital hallways.

• Personal toiletries, such as a toothbrush, toothpaste, and unscented lotion.

• Containers for glasses, contact lenses, or dentures.

• All current medications in their original bottles.

For the home recovery supply kit, it is essential to have the following on hand ³:

• Dilation Supplies: A set of vaginal dilators (as recommended by the surgeon), a large supply of water-based sterile lubricant (avoid silicone-based), and liquid antibacterial soap for cleaning the dilators.

• Hygiene Supplies: A large supply of unscented maxi pads or pantiliners, mild and unscented soap for showering (like Dove or Cetaphil), a douching kit (a bag is often preferred for gentle pressure), and disposable absorbent bed pads (”chucks”) to protect sheets.

• Comfort Items: A donut-shaped or hemorrhoid pillow to relieve pressure when sitting, a handheld mirror for dilation and self-inspection, comfortable cotton underwear, and protein shakes for easy nutrition.

By taking the time to meticulously plan these logistical details, an individual can create a healing environment that is safe, supportive, and low-stress, allowing them to focus their energy entirely on the important work of recovery.

The Surgery and Your Recovery

Your Surgical Day and Hospital Stay

The day of surgery is a momentous occasion, the culmination of months or even years of preparation. While it is natural to feel a mix of excitement and anxiety, understanding the process can help demystify the experience and provide a sense of calm and readiness. The immediate post-operative period is a time of managed dependency, where the medical team creates a controlled environment to give the body the best possible start to the healing process.

The Day of Surgery: What to Expect

Upon arrival at the hospital, the pre-operative process will begin. This may involve washing with a special antiseptic skin cleanser to reduce the risk of infection.¹⁰ The individual will meet with the members of their surgical team, including the anesthesiologist, who will review their medical history and explain the anesthesia plan.¹⁵ An intravenous (IV) line will be started to administer fluids and medications, including antibiotics to prevent infection.¹⁰ The surgery itself is performed under general anesthesia, meaning the individual will be completely asleep and will not feel any pain during the procedure.¹⁰ Vaginoplasty is a complex and meticulous operation that typically takes between five and eight hours to complete.¹⁵

Inside the Operating Room: A Gentle Overview

While the patient is comfortably asleep, the surgical team works to transform and reconstruct the genital tissues. The procedure is performed through incisions in the perineum, the area between the genitals and the anus.¹⁶ A thin, flexible tube called a urinary catheter is placed into the bladder to drain urine during and after the surgery.¹⁰ The surgeon then proceeds with the chosen technique to create the neovagina, neoclitoris, and vulva.

At the conclusion of the surgery, the team implements several measures to aid in initial healing.. Small tubes called surgical drains may be placed near the incisions to gently remove any excess fluid that can accumulate, which helps to reduce swelling and promote healing.¹⁰ The newly created neovagina is then filled with a special surgical dressing known as vaginal packing or a stenting device. This packing serves a crucial purpose: it applies gentle pressure to control bleeding and, most importantly, provides internal support to the delicate new tissues, holding them in place and ensuring the vaginal canal maintains its intended shape and depth as it begins to heal.¹⁰

Your Hospital Stay: The First Few Days of Healing

Following surgery, the individual will be moved to a recovery room for monitoring before being transferred to their hospital room. The typical inpatient hospital stay after vaginoplasty is about five to seven days.¹⁰ This initial period is dedicated to ensuring a safe and stable start to the recovery process. The primary focuses of care are pain management, wound care, and close observation for any immediate post-operative complications.⁶

For the first few days, strict bed rest is usually required to allow the tissues to begin healing without any strain.¹⁶ The nursing team provides comprehensive care, assisting with all needs and administering pain medication to ensure comfort.⁶ The urinary catheter, surgical drains, and vaginal packing remain in place during this time, managing bodily functions externally and supporting the surgical site internally. This allows the body to dedicate all of its energy to the critical first phase of healing. Just before discharge from the hospital, or sometimes at the first follow-up appointment a few days later, the medical team will remove the vaginal packing, catheter, and drains.¹⁰ This is a significant milestone that marks the transition from the hospital-managed phase of recovery to the at-home, self-care phase of the journey.

The First Weeks at Home: Healing and Self-Care

The return home from the hospital marks the beginning of a new and crucial phase of the recovery journey. This period, typically lasting six to eight weeks, requires a profound commitment to rest, patience, and diligent self-care. It is a time for both physical and emotional healing, and understanding what to expect can help navigate this transformative stage with confidence and self-compassion.

The Recovery Timeline: Patience is Key

It is essential to have realistic expectations about the recovery timeline. Vaginoplasty is a major surgery, and the body needs ample time to heal. Most people require about six to eight weeks off from work or school.³ During this time, physical activity must be strictly limited. Strenuous activities, heavy lifting (anything over 15 pounds), swimming, cycling, and horseback riding are prohibited for at least six to eight weeks, and sometimes longer, based on the surgeon’s specific instructions.¹⁰

The healing process is gradual and unique to each individual.¹⁴ The final aesthetic and functional results will not be apparent for many months, as it can take up to a year for all the internal and external swelling to fully subside and for the tissues to soften and settle into their final form.¹ Patience is perhaps the most important tool during this period.

Managing Your Body’s Response: Pain, Swelling, and Discharge

In the first few weeks at home, it is normal to experience some level of pain, significant swelling, and bruising in the genital area.¹ This is a natural part of the body’s inflammatory response to surgery. Pain will be managed with prescribed medications, which should be taken as directed. It is also common for surgeons to prescribe stool softeners and recommend a high-fiber diet to prevent constipation and avoid straining during bowel movements, which could put pressure on the healing tissues.¹⁶

Initially, a brownish-yellow discharge can be expected for the first four to six weeks, and light bleeding or spotting may also occur for up to eight weeks.¹¹ Light bleeding or spotting can also occur for up to eight weeks.¹¹ This is part of the healing process as the internal tissues mend. Regular, gentle douching as instructed by the surgeon will help to keep the neovagina clean and manage this discharge.

Wound and Hygiene Care

Following the surgeon’s instructions for wound care and hygiene is critical to prevent infection and promote proper healing.¹ Showering is usually permitted a day or two after returning home or after the first post-operative visit. When showering, it is important to let warm, soapy water run gently over the genital area without scrubbing. Afterward, the incisions should be carefully patted dry with a clean towel.¹¹ Soaking in a bathtub is not allowed for at least eight weeks to protect the healing incisions from bacteria.¹⁰

As mentioned, douching will become a regular part of the hygiene routine. The surgeon will provide specific instructions on how often to douche and what to use. Typically, this involves gently rinsing the neovagina with a solution of warm water and a small amount of mild, unscented soap.¹⁰ This helps to flush out discharge, residual lubricant from dilation, and any dissolving stitches, keeping the canal clean and healthy.

The Emotional Recovery Journey

The recovery period is as much an emotional process as it is a physical one. It is completely normal to experience a wide and sometimes contradictory range of feelings.²⁶ Moments of profound joy, relief, and excitement may be interspersed with periods of anxiety, frustration, vulnerability, or even a temporary phase of sadness often termed ‘post-operative depression’.¹⁹

It is vital to understand that experiencing these challenging emotions does not mean the surgery was a mistake or that something is wrong.²⁶ It is a natural psychological response to a major life event, the effects of anesthesia, the discomfort of physical healing, and the temporary limitations on one’s independence. Some individuals may also experience a temporary resurgence of gender dysphoria as they navigate the swelling and bruising of the healing process, waiting for the final results to emerge.¹⁹

The connection between physical and emotional healing is undeniable; physical discomfort can amplify feelings of sadness, while emotional stress can impede the body’s ability to heal. Therefore, actively caring for one’s mental health is a non-negotiable part of the recovery plan. Strategies for emotional self-care include:

• Practicing Self-Compassion: Giving oneself permission to rest, to feel all emotions without judgment, and to acknowledge the magnitude of the journey undertaken.²⁶

• Staying Connected: Leaning on one’s established support system of friends, family, and peers. Talking about feelings can reduce isolation and provide much-needed validation.²⁰

• Engaging in Gentle, Soothing Activities: As energy allows, engaging in activities like journaling, mindfulness or meditation, listening to favorite music, or watching comforting movies can help lift one’s spirits.²⁶

• Seeking Professional Support: Continuing to meet with a therapist who is familiar with post-surgical experiences can provide essential tools and a safe space to process the complex emotions of recovery.¹⁹

Prioritizing emotional well-being with the same diligence as wound care is a fundamental component of a holistic and successful recovery.

The Art and Science of Dilation

Of all the post-operative responsibilities, none is more critical to the long-term success of a full-depth vaginoplasty than dilation. This practice is not merely a suggestion but an absolute necessity for maintaining the results of the surgery. While it requires discipline and a lifelong commitment, it is helpful to reframe dilation not as a clinical chore, but as a profound and ongoing act of self-care and embodiment. It is the process through which an individual actively co-creates and maintains their new body, long after the surgeon’s work is complete.

Why Dilation is Essential

To explain clearly and without alarm, the biological principle behind dilation is straightforward: after a new space like a neovagina is created, the body’s powerful natural healing response is to contract and close that space, much like a piercing will close if an earring is not kept in place.²⁷ Dilation is the consistent and methodical use of medical instruments called vaginal dilators to gently stretch the tissues of the neovagina. This regular stretching counteracts the body’s tendency to contract, thereby preserving the depth and width of the vaginal canal that was achieved in surgery.²⁷

Without diligent dilation, the neovagina will inevitably become shorter and narrower, a condition known as vaginal stenosis.¹⁶ Stenosis can make penetrative sex difficult or impossible and can complicate routine pelvic exams. Therefore, adherence to the dilation schedule prescribed by the surgeon is the single most important factor in preventing this complication and ensuring the functional success of the surgery.⁷ It is also important to understand that while sexual activity involving penetration is a healthy and affirming part of life after recovery, it is not a substitute for the structured, methodical stretching provided by dilators, especially during the crucial first year of healing.¹⁶

Your Dilation Schedule: A General Guide

The first dilation session is typically guided by a nurse or physical therapist before discharge from the hospital.²⁸ While each surgeon’s protocol can vary slightly, the general schedule is phased, with the frequency and duration of dilation changing as the neovagina heals and stabilizes over time.¹¹

• Acute Phase (Months 0-3): This is the most intensive period. Dilation is typically required three times every day. Each session lasts between 10 and 30 minutes.¹¹ The focus during this phase is on maintaining the full depth of the canal as the primary healing takes place.

• Stabilization Phase (Months 3-12): As the tissues become more stable and scar tissue remodels, the frequency of dilation can be gradually reduced. This might taper down to once or twice a day for the next few months, and then further to three to five times per week towards the end of the first year.¹¹ During this phase, the surgeon will likely give clearance to begin progressing to larger-sized dilators to also maintain the width, or girth, of the canal.

• Maintenance Phase (Lifelong): After the one-year mark, the neovaginal tissues are generally stable. However, to prevent gradual stenosis over time, a maintenance dilation schedule is required for life. This typically involves dilating one to three times per week indefinitely.²⁹ If at any point the vagina begins to feel tight, the frequency should be temporarily increased until dilation becomes easier again.²⁹

The Dilation Process: A Step-by-Step Guide

Learning the technique of dilation can feel intimidating at first, but with practice, it becomes a manageable routine. The process should always be approached with gentleness and patience.¹¹

1. Preparation: Begin by washing both your hands and the dilator(s) you will be using with warm water and antibacterial soap. Rinse the dilator thoroughly and dry it with a clean paper towel. Find a comfortable and private space where you can relax. Most people find it easiest to dilate while lying on their back with their knees bent and legs apart. Placing a pillow under the upper body or knees can help with comfort.

2. Lubrication: Apply a generous amount of a sterile, water-based lubricant (such as K-Y Jelly or Surgilube) to the tip and shaft of the dilator. It is essential to only use water-based products, as silicone-based lubricants can damage silicone dilators and may not be ideal for the neovaginal tissues.

3. Insertion: Using a handheld mirror can be very helpful, especially in the beginning, to visualize the vaginal opening. Take a few deep, relaxing breaths to release any tension in the pelvic floor muscles. Gently insert the lubricated dilator into the neovagina. The initial angle of insertion should be slightly downward, toward the small of the back. Once the dilator is partially in, the angle should be adjusted to be more horizontal or slightly upward toward the belly button to follow the natural curve of the canal.

4. Dilation: Continue to gently guide the dilator inward until it reaches the full depth of the canal, stopping when you feel moderate pressure or resistance. There should not be sharp pain. Hold the dilator in place for the prescribed duration. Some surgeons may recommend gentle in-and-out or rotating movements during the session.

5. Progression: The dilation journey begins with the smallest-sized dilator provided by the surgeon. After a period of consistent use (often the first month), once the tissues have adapted, the surgeon will give instructions to begin incorporating the next size up. The typical method is to start each session with the smaller dilator for a few minutes to gently stretch the opening, and then switch to the largest comfortable dilator for the remainder of the session.²⁸

A Trauma-Informed Approach to Dilation

It is important to acknowledge that the act of dilation can be emotionally and psychologically complex. For some, the physical discomfort, combined with the clinical nature of the task, can be challenging. For individuals with a history of body dysphoria or trauma related to their genitals, the process can sometimes feel triggering.²⁷

Adopting a trauma-informed and mindful approach can transform the experience. Rather than viewing it as a mechanical task, dilation can be reclaimed as a gentle, affirming act of tending to one’s own body.²⁷ Creating a calming and positive environment can make a significant difference. This might involve lighting a candle, playing soothing music, or practicing deep breathing and grounding exercises before and during the session. It is an act of resilience and reconnection with a part of the body that is now in alignment with the self. If feelings of anxiety or distress arise, it is okay to pause, breathe, and restart when ready. Consistent communication with the surgical team or a pelvic floor physical therapist about any significant pain, resistance, or emotional difficulty is crucial for receiving the support and guidance needed to make this lifelong practice a sustainable and positive part of one’s self-care routine.

Life After Vaginoplasty

Your New Body: Function, Sensation, and Sexuality

The journey of vaginoplasty culminates in a new relationship with one’s body—one that is, for many, more authentic, comfortable, and joyful. Life after surgery involves adapting to new physical functions, exploring new pathways to sexual pleasure, and experiencing the profound psychological benefits of living in a body that aligns with one’s gender identity.

Long-Term Outcomes and Satisfaction

The available body of research consistently demonstrates that vaginoplasty is a highly beneficial and successful procedure in terms of patient satisfaction. Overwhelmingly, individuals who undergo the surgery report that it has a positive impact on their well-being. Large-scale reviews and specific studies show that the vast majority of patients are satisfied with the outcome.³¹ For instance, one study found that at a mean of five years post-surgery, over 90% of individuals reported that their expectations for life as a woman had been fulfilled, and over 85% saw themselves as women.³² Satisfaction with both the aesthetic appearance and the functional results of the surgery is consistently high.³⁴ These outcomes underscore the effectiveness of vaginoplasty as a crucial medical treatment for gender dysphoria, leading to significant improvements in quality of life, self-esteem, and overall happiness.³¹

Changes in Urination

During vaginoplasty, the urethra is shortened and its opening is repositioned to a location typical for a vulva.⁴ This anatomical change means that after surgery, individuals will need to sit to urinate. It may also alter the physical sensations associated with urination, such as the feeling of a full bladder or the urge to go.⁹ One important long-term consideration is an increased susceptibility to urinary tract infections (UTIs), as the shorter urethra provides a quicker path for bacteria to reach the bladder. It is therefore important to be aware of the symptoms of a UTI—such as a frequent or urgent need to urinate, a burning sensation during urination, or cloudy urine—and to seek prompt medical attention if they occur.³ Urinating after dilation or sexual activity can help to reduce this risk.²⁹

Sensation and Sexuality

One of the primary goals of modern vaginoplasty techniques is the creation of a sexually sensitive vulva. The neoclitoris, meticulously crafted from the nerve-rich tissue of the glans penis, is designed to be the focal point of erotic sensation and is capable of producing orgasm.³ Studies confirm that patient satisfaction with the sensitivity of the neoclitoris is generally very high.³⁶

The neovaginal canal itself has a different sensory profile. While it is typically sensitive to pressure, which can be pleasurable during intercourse, it generally does not have the same capacity for fine-touch sensation as the neoclitoris.⁹ The journey of discovering one’s new sexual response is a process of exploration and learning. It involves becoming reacquainted with the body and discovering what types of touch and stimulation lead to arousal and pleasure.

Research into post-operative sexuality reveals overwhelmingly positive experiences. A significant number of individuals report that their orgasms after surgery are more intense and pleasurable than those they experienced before.³⁶ While some may find that orgasms are achieved less frequently than before, the overall quality of the experience is often enhanced.³⁷ The pleasure derived from sexual activity has been found to correlate strongly with the sensitivity of the neoclitoris, more so than with the depth of the vaginal canal.³⁶ This finding is significant, as it highlights that the quality of sexual satisfaction is more dependent on the meticulous creation of a sensitive clitoris than on achieving maximum vaginal depth. This is a crucial point of discussion to have with a surgeon when defining personal goals for the surgery. Even in cases where a neovagina is self-lubricating, such as with PPT or colovaginoplasty, the use of additional water-based lubricant during sexual activity is often recommended for comfort and pleasure.⁹

Revision Surgeries

In some cases, a secondary or revision surgery may be desired or necessary to refine the results. This is not necessarily a sign of a complication, but rather a planned part of the process for some surgical approaches. For example, some surgeons perform the initial vaginoplasty and then, several months later after the initial swelling has resolved, perform a second procedure called a labiaplasty to create or further define the labia minora and clitoral hood.⁷ Other revisions might be sought to address aesthetic concerns or to correct minor functional issues that arise during healing.

Navigating Potential Complications

Every major surgical procedure carries inherent risks, and it is essential for anyone considering vaginoplasty to have a clear and realistic understanding of the potential complications. This knowledge is not meant to cause alarm, but to empower individuals to make a fully informed decision and to be vigilant partners in their own post-operative care. Early recognition of a potential issue is key to successful management and preventing more serious problems.

Understanding General Surgical Risks

Like any operation performed under general anesthesia, vaginoplasty carries a small risk of complications such as adverse reactions to anesthesia, bleeding, the formation of blood clots in the legs or lungs (venous thromboembolism), and post-operative infection.¹ Adhering to pre-operative instructions, such as stopping hormones and nicotine, is a critical step in minimizing these risks.

Complications Specific to Vaginoplasty

Studies show that while minor complications are relatively common, serious complications are rare.³⁸ The overall complication rate in a large population-based cohort was found to be around 25%, with about half of those requiring some form of surgical intervention to correct; it is also noteworthy that complication rates can vary significantly depending on the surgical facility and the experience of the surgical team.³⁸

Potential complications specific to vaginoplasty include ⁷:

• Wound Healing Issues: Minor wound separation (dehiscence) or delayed healing can occur, particularly in areas of high tension like the vaginal opening. These can often be managed with local wound care.

• Tissue Necrosis: This refers to the death of some of the skin or tissue used for the reconstruction, often due to compromised blood supply. Minor cases can heal on their own, but more significant tissue loss may require surgical debridement and repair.⁴⁰

• Vaginal Stenosis: This is the narrowing and shortening of the vaginal canal. As discussed extensively, it is the most common long-term functional complication and is almost entirely preventable with strict adherence to a lifelong dilation regimen.¹⁶

• Fistulas: A fistula is a rare but serious complication involving an abnormal opening between the neovagina and a nearby organ. A rectovaginal fistula is a connection to the rectum, which can cause stool to pass into the vagina. A urethrovaginal fistula is a connection to the urethra, which can cause urine to leak into the vagina. Both typically require surgical repair.¹⁶

• Urinary Issues: Some individuals may experience a temporarily altered or spraying urinary stream as the urethra heals. Stenosis (narrowing) of the new urethral opening can also occur and may require a minor procedure to correct.⁷

• Prolapse: In rare cases, the neovaginal lining can prolapse, or descend from its position.

• Hematoma: This is a collection of blood under the skin that causes significant bruising and swelling. While some bruising is normal, a large hematoma may need to be surgically drained.¹⁶

A Note on Granulation Tissue and Silver Nitrate Treatment

During the healing process, it is common for the body to form new connective tissue and blood vessels at the surgical site. Sometimes, the body’s healing response can be a bit too enthusiastic, leading to an overgrowth of this tissue, known as hypergranulation or simply granulation tissue. This is one of the most common minor complications after vaginoplasty, often appearing as small, pink or beefy-red, shiny bumps that look a bit like skin tags or the surface of a raspberry. This tissue is most likely to form along suture lines, especially inside the neovagina or around the vulva, where the environment is moist and subject to friction from dilation.

While granulation tissue itself is not dangerous, it can be a source of frustration and concern during recovery. Because it is rich in blood vessels, it is very delicate and bleeds easily. This can lead to spotting or bleeding on your dilator or underwear, which can be alarming. It can also cause discomfort or pain during dilation. If left untreated, extensive granulation tissue can interfere with proper healing, lead to scarring, and in some cases, contribute to a loss of vaginal depth or width.

Fortunately, granulation tissue is highly treatable. The most common and effective treatment is the application of silver nitrate. Silver nitrate is a chemical compound applied using a small stick that cauterizes, or burns, the excess tissue. This process is not as scary as it sounds; it causes the granulation tissue to dry up and fall off over a day or two, allowing the healthy skin underneath to heal properly. The treatment may cause the area to turn gray or black temporarily, which is a normal reaction. While the application can be mildly uncomfortable, it is generally well-tolerated. Often, several treatments, typically once a week for a few weeks, are needed to fully resolve the issue.

In addition to silver nitrate, your surgeon or a knowledgeable local provider may recommend other approaches. These can include topical steroid creams or medical-grade honey, which can be applied to the tip of your dilator to treat granulation tissue inside the neovagina. In very rare and persistent cases, a minor surgical procedure to excise the tissue may be considered, but this is seldom necessary. If you notice persistent bleeding with dilation or see tissue that matches this description, it is important to reach out to your surgical team or a local gynecologist. Proactive treatment is the best way to ensure a smooth healing process and protect the long-term results of your surgery.

When to Seek Immediate Medical Attention

Being an active participant in one’s recovery means knowing the warning signs that require prompt communication with the surgical team or a visit to an emergency room. These signs include ⁵:

• Signs of a serious infection, such as a fever, chills, or pus draining from an incision.

• Excessive or uncontrolled bleeding from the surgical site.

• Severe pain that is not managed by the prescribed pain medication.

• Symptoms of a blood clot, which include significant, painful swelling in one calf or leg, shortness of breath, or sudden chest pain.

• An inability to keep food or liquids down due to persistent nausea or vomiting.

By being aware of these potential complications and their warning signs, individuals can ensure that any issues are addressed quickly and effectively, leading to a safer and smoother recovery.

A Lifelong Journey of Affirmation

The Long-Term Impact on Well-Being

The journey through vaginoplasty is an intensive one, demanding physical endurance, emotional resilience, and a deep commitment to long-term self-care. Yet, for the vast majority of individuals who choose this path, the destination is one of profound affirmation and improved quality of life. The scholarly literature provides robust evidence that gender transition, including gender-affirming surgeries like vaginoplasty, is an effective and beneficial treatment for gender dysphoria.³¹

The positive outcomes are not merely anecdotal. Systematic reviews of peer-reviewed research show a strong consensus that transition improves the overall well-being of transgender people. The benefits documented in these studies are wide-ranging and include significant reductions in anxiety, depression, and suicidality, alongside increases in self-esteem, relationship satisfaction, and overall life satisfaction.³¹ Regrets following surgery are exceedingly rare and have become even more so as surgical techniques have advanced and social support has improved.³¹

Embracing Your New Reality

Vaginoplasty is more than a physical alteration; it is a reclamation of self. It is a powerful statement of identity and a courageous step toward living a life of authenticity and congruence. The conclusion of the surgical recovery period is not an end, but a beginning—the start of a new chapter of embodiment, exploration, and self-acceptance.

The journey requires ongoing partnership with one’s body. This includes the lifelong physical practice of maintenance dilation, a ritual of self-care that preserves the functional results of the surgery.¹⁶ It also includes the ongoing emotional and psychological work of integrating this new physical reality into one’s life, exploring intimacy and sexuality, and continuing to nurture one’s mental health through therapy, community, and connection.²⁰

The path to and through vaginoplasty is a testament to the resilience of the human spirit. It is a journey of becoming, undertaken with hope and determination. The final message of this guide is one of celebration and empowerment—a tribute to the courage it takes to pursue one’s truth and to create a life in which the body and the self can finally, peacefully, coexist.

Next Steps: A Journey Inward

If you’re reading this, you are likely navigating one of the most personal and significant aspects of your transition journey: deciding which gender-affirming surgical path to pursue. At this point in the book, we have already laid out all of the options for those assigned male at birth (AMAB). It’s a path filled with possibilities, emotions, and big decisions. My most important piece of advice is this: be gentle with yourself. There is no right or wrong timeline, and there is no single “correct” choice. There is only what is right for you.

You’ve already done the hard work of learning about the different surgical options available, from nulloplasty to various forms of vaginoplasty and vulvoplasty. Now comes the internal work. The following questions are not a test. They are gentle prompts designed to help you listen to your own heart, body, and hopes for the future. Take your time with them. Sit with them. Journal about them. Talk them over with a trusted friend or therapist. There is no rush to find an answer; allow yourself weeks, months, or even years to explore these feelings. The goal is not to force a decision, but to wait patiently until one path forward feels true and resonates more deeply than the others. This is your journey, and your inner wisdom is your most valuable guide.

Section 1: Connecting with Your Body & Feelings of Dysphoria/Euphoria

Before focusing on a specific surgical outcome, let’s start with how you feel right now. Understanding your relationship with your body is the foundation for any decision you make.

• When you think about your current anatomy, what specific feelings come up? Can you name them? (e.g., anxiety, neutrality, frustration, detachment).

• What parts of your current anatomy, if any, cause you the most distress or dysphoria on a daily basis? What does that distress feel like in your body and mind?

• Are there times when you feel comfortable or even euphoric in your body as it is now? What are those moments like?

• Imagine yourself in the future, having had some form of surgery. What is the primary feeling you hope to achieve? Is it relief? Joy? Congruence? Gender euphoria? Peace?

• If you could wave a magic wand and change how you feel about your body without changing its physical form, would that be enough for you? Why or why not?

Section 2: Envisioning Your Daily Life & Functionality

Surgery isn’t just about a moment in time; it’s about how you will live in your body every day afterward. Let’s explore what you want your future to look and feel like.

• Urination: How important is it for you to sit to urinate? Does the idea of this bring you comfort, affirmation, or a sense of rightness? Or is it not a significant factor for you?

• Sexual Function & Intimacy: This is a deeply personal area. Be honest with yourself, without judgment.

  • What role does sex and intimacy play in your life now, and what role would you like it to play in the future?

  • Do you desire the potential for penetrative sex with a vaginal canal? How important is that specific experience to you?

  • Is sensation (specifically, the potential for orgasmic pleasure) a primary goal? Are you more focused on clitoral sensation, or is the idea of internal sensation (like from a vaginal canal) also appealing?

  • How do you feel about the idea of self-lubrication versus using manual lubrication?

  • How important is it to you to preserve the parts of your anatomy that are currently capable of erection? What feelings come up when you consider removing them?

• Daily Physicality: Think about your life beyond intimacy.

  • How do you want to feel when you’re wearing different kinds of clothes (like swimwear, leggings, or underwear)?

  • What about non-sexual physical activities? How do you want to feel when you’re walking, running, sitting, or just existing in the world?

  • Does the idea of “tucking” cause you distress? Is a primary goal to eliminate the need to tuck?

Section 3: Exploring Your Aesthetic Goals

What feels and looks right to you is uniquely yours. Aesthetics are a valid and important part of gender affirmation.

• When you imagine your ideal vulva, what do you see? Is it important for it to look a certain way?

• Does the idea of having a vaginal opening feel symbolically or aesthetically important to you, even if you don’t plan to use it for penetrative sex?

• Conversely, does the idea of a smooth, uninterrupted perineum (as with nulloplasty) bring you a sense of peace, comfort, or aesthetic satisfaction?

• Look at diagrams or post-operative photos (if you feel comfortable doing so). Which results give you a feeling of “yes, that’s it”? What about them resonates with you?

Section 4: Practical & Logistical Considerations

These questions are about the “how.” The journey involves more than just the surgery itself.

• Recovery & Maintenance:

  • How do you feel about a long and intensive recovery period (months to a year or more for some procedures)? Do you have the time and stability in your life to accommodate that?

  • Are you prepared for the lifelong commitment of dilation required for maintaining a vaginal canal? How does the idea of incorporating that into your routine feel?

  • How do you feel about a shorter, less complex recovery with fewer long-term maintenance needs (as with vulvoplasty or nulloplasty)?

• Risk & Health:

  • Every surgery has risks. How comfortable are you with the potential for complications associated with more complex procedures versus less complex ones?

  • Do you have any underlying health conditions that might make a longer, more invasive surgery a greater risk?

• Support System:

  • Who will be there to take care of you during your recovery? Is your support system solid and reliable for both physical and emotional needs?

  • How will you handle your financial needs and work obligations during your time off?

A Final Thought

As you sit with these questions, you may not find immediate, clear answers. That is more than okay. Sometimes, the answer is “I’m not sure yet,” and that is a perfectly valid place to be. This is a process of discovery. Your answers might change over time. The goal isn’t to rush to a finish line, but to walk your path with intention, self-compassion, and honesty. You deserve to feel at home in your body, and you have the strength and wisdom to find out what “home” means to you.

Subscribe now

Works cited

1. MTF Surgery | IM GENDER, accessed October 9, 2025, https://imgender.com/en/mtf-surgery/

2. Vaginoplasty Techniques: Which One to Choose? | IM GENDER, accessed October 9, 2025, https://imgender.com/en/vaginoplasty-technique/

3. Vaginoplasty Guidebook | Women’s College Hospital, accessed October 9, 2025, https://www.womenscollegehospital.ca/wp-content/uploads/2022/09/VaginoplastyGuidebook.pdf

4. Vaginoplasty and Vulvoplasty | OHSU, accessed October 9, 2025, https://www.ohsu.edu/transgender-health/vaginoplasty-and-vulvoplasty

5. Phallus preserving vulvovaginoplasty, accessed October 9, 2025, https://irp.cdn-website.com/8ed2c9d8/files/uploaded/11733882022.pdf

6. Sex Reassignment Surgery (SRS) | IM GENDER, accessed October 9, 2025, https://imgender.com/en/sex-reassignment-surgery/

7. “Fem” Bottom Surgery - Gender Confirmation, accessed October 9, 2025, https://www.genderconfirmation.com/vaginoplasty-transgender-vagina/

8. Phallus Preserving Vaginoplasty | ART Surgical, accessed October 9, 2025, https://www.artsurgical.net/phallus-preserving-vaginoplasty

9. Adding to Your Bottom Line: Phallus-Preserving VulvoVaginoplasty ..., accessed October 9, 2025, https://queerdoc.com/phallus-preserving-vaginoplasty/

10. Vaginoplasty: Surgery and After Care | UMass Memorial Health, accessed October 9, 2025, https://www.ummhealth.org/health-library/vaginoplasty-surgery-and-after-care

11. Vaginoplasty procedures, complications and aftercare | Gender ..., accessed October 9, 2025, https://transcare.ucsf.edu/guidelines/vaginoplasty

12. Use of right colon vaginoplasty in gender affirming surgery ..., accessed October 9, 2025, https://www.researchgate.net/publication/344667500_Use_of_right_colon_vaginoplasty_in_gender_affirming_surgery_proposed_advantages_review_of_technique_and_outcomes

13. Use of right colon vaginoplasty in gender ... - eScholarship.org, accessed October 9, 2025, https://escholarship.org/content/qt7pv6h1h2/qt7pv6h1h2.pdf

14. What to Expect With Penile-Preserving Vaginoplasty | ART Surgical, accessed October 9, 2025, https://www.artsurgical.net/blog/what-to-expect-with-penile-preserving-vaginoplasty

15. Feminine/Non-Binary Adult Gender-Affirming Surgery - Temple Health, accessed October 9, 2025, https://www.templehealth.org/services/adult-gender-affirming-healthcare/surgeries-procedures/feminine-non-binary

16. Vaginoplasty - Brigham and Women’s Faulkner Hospital, accessed October 9, 2025, https://www.brighamandwomensfaulkner.org/programs-and-services/urology/vaginoplasty

17. Everything you Need to Get Bottom Surgery, accessed October 9, 2025, https://www.genderconfirmation.com/how-to-get-bottom-surgery/

18. Vulvaplasty Guidebook - Women’s College Hospital, accessed October 9, 2025, https://www.womenscollegehospital.ca/wp-content/uploads/2025/05/Vulvaplasty-Guidebook-May-2025-final.pdf

19. Dysphoria After Surgery: Coping Tips for Emotional Wellness | ART ..., accessed October 9, 2025, https://www.artsurgical.net/blog/dysphoria-after-surgery-coping-tips-for-emotional-wellness

20. Supporting the Transgender Community | ART Surgical, accessed October 9, 2025, https://www.artsurgical.net/transgender-community

21. Vaginoplasty Post-Op Supplies Checklist - Temple Health, accessed October 9, 2025, https://www.templehealth.org/sites/default/files/2024-06/check-list-vaginoplasty-supplies.pdf

22. Preparing Your Home for Gender-Affirming Surgery Recovery - ART Surgical, accessed October 9, 2025, https://www.artsurgical.net/blog/home-gender-affirming-surgery-recovery

23. Preparing for Your Hospital Stay - VulvoVaginoplasty - QueerDoc, accessed October 9, 2025, https://queerdoc.com/media/2023/08/Preparing-for-Your-Hospital-Stay-VulvoVaginoplasty.pdf

24. Post-Op Instructions for Urogynecology - Dr.Vitasna.com, accessed October 9, 2025, https://www.drvitasna.com/post-op-instructions-for-urogynecology/

25. Transgender Vaginoplasty San Francisco - ART Surgical, accessed October 9, 2025, https://www.artsurgical.net/transgender-vaginoplasty

26. Emotional Healing After Transgender Surgery | ART Surgical, accessed October 9, 2025, https://www.artsurgical.net/blog/emotional-healing-after-transgender-surgery

27. Your Guide to Dilation After Vaginoplasty: Healing, Pleasure ..., accessed October 9, 2025, https://www.vuvatech.com/blogs/care/your-guide-to-dilation-after-vaginoplasty-healing-pleasure-reclaiming-your-body

28. Vaginal Dilatation: How To Do It | IM GENDER, accessed October 9, 2025, https://imgender.com/en/vaginal-dilatation/

29. DILATING INSTRUCTIONS We recommend Soul Source dilators ..., accessed October 9, 2025, https://www.womenscollegehospital.ca/wp-content/uploads/2022/09/DilatingInstructions.pdf

30. How Long Do You Have to Dilate After MTF? A Complete Guide to Dilation - Vuvatech, accessed October 9, 2025, https://www.vuvatech.com/blogs/care/how-long-do-you-have-to-dilate-after-mtf-a-complete-guide-to-dilation-after-vaginoplasty

31. What does the scholarly research say about the ... - What We Know, accessed October 9, 2025, https://whatweknow.inequality.cornell.edu/topics/lgbt-equality/what-does-the-scholarly-research-say-about-the-well-being-of-transgender-people/

32. Satisfaction With Male-to-Female Gender Reassignment Surgery: Results of a Retrospective Analysis - PubMed Central, accessed October 9, 2025, https://pmc.ncbi.nlm.nih.gov/articles/PMC4261554/

33. Satisfaction with male-to-female gender reassignment surgery, accessed October 9, 2025, https://pubmed.ncbi.nlm.nih.gov/25487762/

34. [PDF] Satisfaction with male-to-female gender reassignment surgery ..., accessed October 9, 2025, https://www.semanticscholar.org/paper/Satisfaction-with-male-to-female-gender-surgery.-Hess-Neto/5608e4bf3e7d5e682b05eb515744af46eef0b173

35. (PDF) Satisfaction With Male-to-Female Gender Reassignment ..., accessed October 9, 2025, https://www.researchgate.net/publication/270512898_Satisfaction_With_Male-to-Female_Gender_Reassignment_Surgery

36. Sexuality after Male-to-Female Gender Affirmation Surgery - PubMed, accessed October 9, 2025, https://pubmed.ncbi.nlm.nih.gov/29977922/

37. Sexuality after Male-to-Female Gender Affirmation Surgery - PMC - PubMed Central, accessed October 9, 2025, https://pmc.ncbi.nlm.nih.gov/articles/PMC5994261/

38. PD31-08 COMPLICATIONS AFTER GENDER AFFIRMING ..., accessed October 9, 2025, https://www.auajournals.org/doi/10.1097/JU.0000000000002032.08

39. Complications Following Gender-affirming Penile-inversion ... - AAPS, accessed October 9, 2025, https://meeting.aaps1921.org/program/2025/P75.cgi

40. Complications in transgender patients undergoing vaginoplasty ..., accessed October 9, 2025, https://www.scielo.org.mx/article_plus.php?pid=S2444-054X2025000200007&tlng=en&lng=es

41. Granulation Tissue Treatment Following MTF Vaginoplasty Surgery, accessed October 9, 2025, https://keeleemacpheemd.com/granulation-tissue-treatment/

42. HYPERGRANULATION AFTER PENILE INVERSION ..., accessed October 9, 2025, https://www.womenscollegehospital.ca/wp-content/uploads/2022/09/Hypergranulation.pdf

43. Vaginoplasty- Common Complications | Women’s College Hospital, accessed October 9, 2025, https://www.womenscollegehospital.ca/wp-content/uploads/2022/09/VaginoplastyCommonComplications.pdf

44. MoZaic Care: Granulation Tissue - YouTube, accessed October 9, 2025 https://www.youtube.com/watch?v=nyGOooFigP4&pp=ygUfTW9aYWljIENhcmU6IEdyYW51bGF0aW9uIFRpc3N1ZQ%3D%3D

45. PART C - GrS Montréal, accessed October 9, 2025, https://www.grsmontreal.com/DATA/TEXTEDOC/Booklet-VAGINOPLASTY-WITHOUT-CAVITY-Part-C.pdf

46. Bottom Surgeries, Wound Healing, and Granulation Tissue ..., accessed October 9, 2025, https://queerdoc.com/10-9-bottom-surgery-granulation-tissue/

47. Vaginal and Perineal Granulation Tissue - The Vagina Whisperer, accessed October 9, 2025, https://thevagwhisperer.com/2021/12/14/perineal-tearing/

48. Hypergranulation management following penile inversion vaginoplasty | Request PDF, accessed October 9, 2025, https://www.researchgate.net/publication/383011418_Hypergranulation_management_following_penile_inversion_vaginoplasty

49. Silver Nitrate: Uses, Tips, Side Effects & More - GoodRx, accessed October 9, 2025, https://www.goodrx.com/silver-nitrate/what-is

50. Silver Nitrate: What You Need To Know - The Wound Pros, accessed October 9, 2025, https://www.thewoundpros.com/post/silver-nitrate-what-you-need-to-know

51. 1212 - Silver nitrate for granulation tissue - Children’s Wisconsin, accessed October 9, 2025, https://childrenswi.org/publications/teaching-sheet/trach-vent/1212-silver-nitrate-for-granulation-tissue

52. Silver Nitrate: Wound Uses, Warnings, Side Effects, Dosage - MedicineNet, accessed October 9, 2025, https://www.medicinenet.com/silver_nitrate/article.htm

Taking this moment to read these words is a big deal. It’s a courageous step on what might be the most significant journey of your life: the journey home to yourself. If you’re here, you’re probably already familiar with the deep, quiet work of figuring out who you are—of untangling the world’s expectations from the truth of your own heart. It’s about learning to listen to that inner voice and finding the bravery to honor what it tells you. This path is yours and yours alone. There’s no map, no timeline, and no right or wrong way to travel it. Every single step is an act of profound self-love.

For many, this journey involves finding new ways to express themselves socially, like trying out new names, pronouns, or styles that feel more like them. For others, it’s about using hormones to bring their body into a new alignment. And for some, particularly those whose authentic self is agender, non-binary, or exists beyond the traditional gender binary, the path may lead toward a specific kind of physical harmony—a feeling of rightness that can be realized through a surgical option known as nullification, or nulloplasty.

Think of this chapter as a quiet, safe space to explore this option. It’s not meant to steer you in any direction, but to serve as a compassionate and clear guide. My purpose here is to offer respectful information, free of judgment and filled with unwavering support, so you have the tools you need to make the choices that are right for you, with confidence and peace of mind.

Understanding Nullification: A Path to Feeling Right

For many, the goal of gender affirmation isn’t about transitioning from one end of the binary to the other. It’s a much more personal search for a sense of peace in a body that truly reflects a neutral or agender identity. This is about more than just alleviating discomfort; it’s about finding a quiet, profound rightness and finally feeling at home in your own skin. For individuals whose gender exists outside the binary, the presence of prominent sex characteristics can feel like a constant, incorrect statement. Nullification offers a powerful, personal way to resolve that conflict, allowing your physical form to align with your internal truth and achieve that deeply sought-after feeling of wholeness.

What is Nullification?

Nullification, which also goes by the names nulloplasty or nullectomy, is a deeply personal form of gender-affirming surgery. Its main purpose is to create a smooth, uninterrupted landscape from the abdomen down through the groin by removing external genitalia.¹ This goal, however, goes far beyond a simple aesthetic preference; it’s about creating a physical form that feels quiet, calm, and unburdened by characteristics that feel wrong. It’s crucial to understand how this is fundamentally different from other forms of “bottom surgery.” While procedures like vaginoplasty or phalloplasty are designed to construct new genitals that align with a binary gender identity, nullification’s philosophy is entirely different. The focus here isn’t on creation or substitution, but on achieving a serene, neutral physical form—a body that is complete without prominent sex characteristics.¹ This choice represents a move away from a framework of “becoming” something else (male or female) and into a space of simply being—allowing the authentic self to be reflected in a body that is finally at peace.

For individuals who identify as non-binary, agender, or otherwise outside the male/female binary, a deep-seated sense that the ones present are an alien and incorrect feature, like a permanent piece of clothing that doesn’t fit. This constant physical dissonance can consume an enormous amount of mental and emotional energy. Nullification directly addresses this by creating a physical form that mirrors their internal sense of self. It is a powerful, self-defining choice for those who feel most authentic in a body that reflects that neutrality. It’s an act of radical self-care that finally allows their outside to be a true and comfortable home, a place where they no longer feel like a guest in their own skin.²

A Little Bit of History

To really get what nullification is about today, it helps to look back. The idea of altering one’s genitals isn’t new—we can see examples going back centuries with eunuchs in various empires, who lived outside the typical male roles of their time.³ But it’s vital to separate these historical practices from modern, affirming care. The motivations were completely different, often tied to social structures, coercion, or religious practice rather than an individual’s own sense of self.

What makes modern nullification surgery possible isn’t just a leap in technology, but a profound and necessary shift in medical ethics and how we understand gender. For much of the 20th century, the medical world saw any deviation from being cisgender as a mental disorder to be “cured.” This mindset meant that early gender-affirming surgeries, while life-saving for many, were designed with a rigid, a binary outcome in mind: to help someone “pass” as either male or female.

The move toward procedures like nulloplasty marks a crucial shift away from this pathologizing view. It reflects the understanding that gender is a spectrum and that gender diversity is a natural and beautiful part of the human experience, not an illness.⁴ This change in thinking has been monumental. It has moved the goal of medical care from “fixing” a problem to supporting a person’s journey toward wholeness. It has empowered patients to be the authors of their own transitions, opening the door for surgical options that aren’t locked into a strict male/female framework. Nulloplasty exists because we’ve finally started making room for identities that live in the vast, authentic spaces beyond the binary, honoring the truth that the goal of affirmation isn’t to fit into a different box, but to finally feel at home in your own skin.⁴

The Nullo Identity and Finding Your People

Choosing to identify with the term “nullo” is a deeply personal and often empowering act. For some, it’s just a way to describe a surgical choice, but for many, it becomes a core part of their identity. It’s a word that gives a name to an internal truth—the feeling of being most yourself in a state of neutrality. To claim this identity is to affirm that your gender isn’t defined by what is present, but perhaps by a peaceful and intentional absence. In a world that often defines people by categories and characteristics, choosing neutrality is a radical statement. It’s about finding a word that resonates with that quiet sense of rightness that comes from aligning your body with your spirit, a feeling that doesn’t require addition, but a gentle subtraction to reveal what was there all along. For many, the label “nullo” is the first time a word has ever felt like a comfortable fit, providing a sense of relief and recognition that can be life-changing in itself.⁵

This journey can feel profoundly lonely at times, especially when so much of the public conversation and even queer representation centers on binary transitions from man to woman or woman to man. You might feel like an outsider even within trans spaces, where the language and shared experiences don’t quite match your own. This specific kind of isolation can be exhausting, making you question the validity of your own path. In this context, the nullo community, while smaller and less visible, is an essential and powerful testament to the true diversity of gender. Finding this community is often a revelation—the moment you realize you are not a puzzle with a missing piece, but a complete picture that others share.

Connecting with others who are on this specific path—whether through dedicated online forums, private social media groups, or platforms like Reddit—provides a unique and profound sense of validation that can’t be found anywhere else. It’s one thing to be accepted by loving allies who try to understand; it’s another thing entirely to be truly seen by someone who has walked the same road, who understands the nuances without you having to explain them. These spaces offer more than just friendship; they are vital, living archives of shared experience. They are hubs for sharing practical knowledge—from detailed surgeon reviews and recovery tips to advice on navigating insurance hurdles—and for finding the deep emotional support needed to navigate such a significant life event. Reading someone else’s story, seeing their post-op joy, understanding their pre-op anxieties, and witnessing their journey reinforces the powerful, grounding truth that you are not alone. It transforms a path that felt solitary and uncertain into a shared journey with a supportive and knowledgeable community beside you.²

Surgical Pathways to a Neutral Form

The specific procedures involved in nullification are tailored to your anatomy and your unique goals. This means the surgeon isn’t following a one-size-fits-all blueprint; they are working with your body’s specific landscape—the amount and type of tissue available, the location of nerves that can be preserved for sensation, and the underlying structures that will support the final result. Your goals are the most important part of the equation, and this conversation goes much deeper than just the final aesthetic. It’s about your vision for yourself, your hopes for sexual sensation, and what will ultimately bring you the most profound sense of peace.

Surgeons who specialize in this care work closely with their patients to develop a personalized plan. Think of this process less like a doctor giving instructions and more like a collaborative design session. You are the expert on your own body and identity; your surgeon brings the technical expertise to translate your internal truth into a physical reality. This partnership, built on trust and open communication, is at the heart of modern gender-affirming care. A good surgeon will listen deeply and ensure that you feel seen, heard, and respected every step of the way, creating a detailed surgical roadmap that feels completely right for you.⁵

For Individuals Assigned Male at Birth (AMAB)

For AMAB individuals, the main goal of a nulloplasty is a smooth, unbroken contour from the pubic bone to the perineum. This is a highly specialized procedure that combines several surgeries in one go. The core steps include a penectomy (removal of the penis), an orchiectomy (removal of the testicles), and a scrotectomy (removal of the scrotal sac).¹ By removing all external structures, the surgeon can then re-sculpt the area to create that smooth, neutral appearance.

One of the most critical parts of this surgery is reconstructing the urethra. It must be shortened and its opening relocated to a position that allows you to urinate comfortably while sitting down. An experienced surgeon will meticulously craft the new opening to prevent it from narrowing over time—a potential complication known as a stricture.⁶ This step is absolutely key for your long-term comfort and quality of life.

While a smooth look is the main objective, the surgical plan is highly customizable, and nowhere is this more true than when it comes to sexual sensation. It’s a common and persistent myth that nullification means an end to genital pleasure. In reality, preserving and even reshaping the potential for orgasm is a central part of the conversation for most people. Because the standards of care in this space are constantly evolving, many of the most experienced nullo surgeons will recommend some of a nerve preservation by default, at least initially. Their primary goal is to minimize the risk of a patient losing the ability to achieve orgasm after surgery.

For those who want to preserve this potential without any external structure, one innovative and increasingly common option is the creation of a “buried” or “hidden” clitoris. This is a meticulous technique where the surgeon carefully preserves a small, highly sensate portion of the glans penis, which is rich with nerve endings. This tissue is then secured deep within the fatty tissue of the mons pubis, just above the new urethral opening. This creates a powerful internal center of erotic sensation that isn’t visible but can be stimulated through direct pressure, touch, or vibration. It’s a way to maintain a primary source of pleasure while achieving the desired smooth aesthetic.⁷

However, this is just one possibility, and the path is not the same for everyone. Some individuals make the informed choice to opt for no glans preservation at all and go on to have vibrant, fulfilling sex lives, discovering new pathways to pleasure that don’t rely on a specific kind of stimulation. The glans doesn’t have to be buried: it can be left exposed too. This underscores the profound diversity of nullo experiences; this is a bespoke surgery customized to the individual. Ultimately, the decisions around sensation are deeply personal. It is absolutely crucial to have a detailed, open, and honest conversation about your questions, fears, and hopes with a well-qualified surgeon who understands the nuances of these procedures.

Alternatively, if you desire a different look that isn’t completely smooth, a “cosmetic” or “no-depth” vaginoplasty is another option. This procedure doesn’t create a vaginal canal but focuses on constructing the external features of a vulva, like labia and a clitoris, using tissue from the penis and scrotum. This results in a dimpled or textured appearance rather than a flat one, offering another valid of gender-affirming anatomy.¹ The choice between these paths is deeply personal and is decided through extensive conversations with your surgeon.

For Individuals Assigned Female at Birth (AFAB)

For AFAB individuals, the surgical path to nullification is also a deeply personalized process, a series of choices you make in partnership with your surgeon to create a result that feels right for you. The journey typically begins with the external structures, focusing on creating that smooth, neutral landscape.

The foundational external procedure is a vulvectomy, which involves removing the labia. This isn’t a single, standard procedure; it’s tailored to your body and your goals. A partial vulvectomy might remove only the labia minora (the inner lips), while a complete or radical vulvectomy would remove both the labia minora and majora (the outer lips). The choice depends entirely on your desired final appearance. The surgeon’s work here is as much artistic as it is technical, carefully sculpting the remaining tissue to create a flat, seamless contour from the mons pubis down to the perineum.

This leads to one of the most significant points of discussion in the entire process: the clitoris. For those seeking the absolute smoothest result possible, with no external prominences, an optional clitorectomy (the complete removal of the clitoris) can be performed. However, this is a deeply personal and permanent choice with profound implications for sexual sensation. Because the clitoris is a primary erogenous zone, its removal will fundamentally alter sexual response. For this reason, many people choose to retain their clitoris, accepting a small prominence in exchange for preserving this vital center of pleasure. This is a crucial, in-depth conversation to have with your surgeon. It’s about openly and honestly weighing your priorities: what is the balance between your aesthetic goals and your functional desires? There is no right or wrong answer, only the one that feels most authentic to you.

Beyond the external work, the internal procedures are more extensive and represent another set of personal choices. A vaginectomy, the complete removal of the vaginal canal, is a core component for many who seek nullification. For these individuals, the presence of the vaginal canal itself is a primary source of gender dysphoria—a constant internal reminder of an anatomy that feels incorrect. Removing it can bring an immense sense of relief and wholeness. It is critical to understand, however, that for crucial health and safety reasons, a vaginectomy necessitates a hysterectomy (the removal of the uterus). Leaving the uterus in place without a vaginal opening would create a closed system, risking dangerous fluid buildup, infection, and making it nearly impossible to screen for or detect future uterine cancers. Therefore, these two procedures are performed together.

Alongside this, many individuals also opt for an oophorectomy, the removal of the ovaries. This procedure often serves a dual purpose. First, it removes the body’s primary source of endogenous hormones like estrogen, which can be a key goal for those seeking a more androgynous hormonal state and an end to menstrual cycles. Second, it completely eliminates the future risk of ovarian cancer. Finally, depending on how the removal of surrounding tissues affects the anatomy, minor urethral modification might be needed. This is typically a small adjustment to ensure the urethral opening is properly positioned so that the urinary stream remains comfortable, directed, and manageable.¹ Each of these decisions—from the shape of the external skin to the removal of internal organs—is a piece of the puzzle. They are all discussed at length during your consultations, allowing you to construct a surgical plan with a team that understands and respects your unique vision for your body.

Preparing for Your Surgical Journey

Embarking on nullification surgery is more than just a medical procedure; it’s a significant life event that requires thoughtful and holistic preparation. This means tending to your mind, your body, and your practical life logistics with intention and care. The foundation for this entire process is building a strong, trusting, and collaborative relationship with your healthcare team. You are not a passive recipient of care; you are the expert on your own life and an active partner in your journey. Feeling safe, seen, and respected by your providers is not a bonus—it’s a requirement.

Choosing the Right Surgeon and Surgical Team

Selecting a board-certified surgeon with extensive, specific experience in gender-affirming care is the single most important decision you will make. It’s essential to find a surgical team that not only possesses exceptional technical skill but also demonstrates a deep, genuine, and respectful understanding of non-binary identities.⁷ This goes far beyond credentials. It’s about how they make you feel. During a consultation, pay attention: Do they use your correct name and pronouns effortlessly? Do they listen to your goals without judgment or trying to steer you toward a more binary outcome? Do they explain complex medical information clearly and patiently, ensuring you understand completely?

Don’t hesitate to “shop around” if you can. Schedule consultations with more than one surgeon to compare their approaches and communication styles. Come prepared with a list of questions, such as:

• “How many nullification or non-binary procedures have you performed?”

• “What is your philosophy on preserving sensation, and what options do you offer?”

• “Can you walk me through the recovery process and what to expect in the first few weeks?”

• “What are the most common complications you see with this procedure?”

Remember that you are interviewing them for a very important role in your life. Centers like the Crane Center for Transgender Surgery and surgeons like Dr. Kenan Celtik are known for specializing in non-binary procedures and offering truly personalized surgical plans, which is the standard of patient-centered care you deserve.⁶,⁹ Ultimately, trust your gut. The right surgeon and team will make you feel confident, supported, and understood.

Getting Ready: Mind, Body, and Logistics

True preparation is a three-pronged approach that goes far beyond just the day of surgery. It’s about creating the best possible conditions for a smooth procedure and a successful recovery. You’ll need to get your mind ready for the emotional journey, your body ready for the physical demands of surgery, and your life ready to accommodate the crucial healing period.

Mental and Emotional Preparation

Your surgical team will likely require one or two letters of support from a licensed mental health professional, in line with WPATH standards of care.¹ It’s helpful to view this requirement not as a bureaucratic hurdle, but as an affirming and protective step. This process ensures you have a dedicated space to reflect, a strong support system in place, and have had the opportunity to thoroughly explore your decision with a professional who understands. This is an excellent time to work with a therapist to set realistic expectations for your recovery. It’s not just about healing physically; there will be emotional ups and downs. Post-operative blues are very real for any major surgery, and having coping strategies for managing pain, boredom, frustration, and even the intense wave of gender euphoria can make a world of difference. Your therapist can help you build this toolkit.

Physical Preparation

Your journey to a healthy recovery starts weeks before you enter the operating room. You will need to be at least 18 years old and undergo a full medical evaluation, including blood work and sometimes an EKG, to ensure you are healthy enough for a major surgery and anesthesia. The informed consent process is a critical part of this. This is not just signing a form; it should be a detailed, ongoing conversation where your surgeon explains all the permanent changes, potential risks (both common and rare), and expected benefits. This is your opportunity to ask every last question until you feel completely confident. Your surgeon will also provide crucial pre-operative instructions. This will include stopping certain medications and supplements (especially blood thinners like aspirin, but also things like fish oil and vitamin E) that can increase bleeding risk. On that same note, it is absolutely critical to openly and honestly discuss any and all substance use with your surgeon and primary care provider before surgery. This is for your safety, as some substances can interfere with anesthesia or healing. Any substance use disorders must also be addressed prior to surgery per WPATH guidelines.

Perhaps most importantly, you will be required to quit all forms of smoking and nicotine—including vaping and patches—at least six weeks before surgery. This is absolutely non-negotiable. Nicotine is a potent vasoconstrictor, meaning it shrinks your blood vessels and severely restricts blood flow. Proper blood flow is essential for bringing oxygen and nutrients to healing tissues, and without it, you face a dramatically increased risk of serious complications like infection and tissue death (necrosis).

The day before surgery, you may need to complete a bowel prep, depending on your surgeon’s instructions. This involves drinking a special solution to clean out your intestines, which is a vital safety measure to minimize infection risk during surgery. While it’s not pleasant, framing it as the final act of self-care before your procedure can help.⁷

Logistical Preparation

Planning for your recovery is one of the most practical and important things you can do to reduce stress and allow your body and mind to heal. It’s critical to remember that your needs will vary based on your specific situation. If you live alone, your needs will be different than if you have a partner. If you have to travel for surgery, your planning will be much more complex. It’s crucial to think through all of the necessary logistical steps to make a successful recovery possible.

Start by arranging for at least three weeks of dedicated, around-the-clock care from a trusted friend, partner, or family member. This person will be your lifeline, helping with everything from managing your medication schedule to preparing food and assisting with basic mobility. Before surgery, prepare your home by creating a comfortable and functional “recovery nest.” This should include plenty of pillows, a small table for water and meds, extra-long charging cables, and any entertainment you might want. If you are traveling, this “nest” will be your hotel room or Airbnb. For my own nulloplasty, I stayed in a hotel room in San Francisco that didn’t even have a microwave. We survived, but the stress of figuring out food every day was significant. If you can, an Airbnb with a kitchen can make a world of difference, allowing your caregiver to prepare simple, nourishing meals.

Think ahead about food. The last thing you’ll want to do is cook. Before surgery, cook and freeze single-serving portions of soups, stews, or casseroles. Stock your pantry and fridge with easy-to-prepare, nutrient-dense foods like protein shakes, yogurt, oatmeal, and fruit. Taking care of all these practical needs beforehand—laundry, cleaning, grocery shopping—will lift a huge weight off your shoulders and allow you to focus completely on your one and only job after surgery: resting and healing. A later chapter will include a full shopping list for post-op care, put together by our community of people who have already gone through this themselves.

Recovery, Healing, and Aftercare

Your body and spirit are about to go through a major event, and recovery is a journey in itself. It deserves time, patience, and dedicated care. It is absolutely crucial to be gentle with yourself. Healing isn’t a race, and there’s no prize for pushing your limits too soon. Listening to your body—fatigue, discomfort, a simple need for quiet—is one of the most important skills you can practice during this time. This period calls for radical self-compassion. This means actively fighting the urge to be a “perfect patient” who never complains or needs anything. Instead, the goal is to be a kind, attentive, and forgiving caretaker to yourself. You will have good days where the pain is manageable and you feel optimistic, and you will have bad days where you feel exhausted, frustrated, and sore. This emotional rollercoaster is not just normal; it’s an expected part of the process. Acknowledge it, and know that each dip is temporary.

It’s also important to be prepared for the phenomenon of “post-operative blues,” a temporary drop in mood that can happen a few days or weeks after any major surgery. Even after a deeply desired and affirming procedure, the physical trauma, the effects of anesthesia, the discomfort, and the sheer boredom of being laid up can lead to feelings of sadness or irritability. This is not a sign that you’ve made a mistake; it’s a physiological and psychological response to a massive undertaking. Knowing this can happen helps you meet those feelings with curiosity instead of fear.

If you can, fiercely resist the immense internal and external pressure to jump right back into the demands of work or school. This isn’t just about letting incisions heal on the surface; it’s about allowing your entire system—physical, mental, and emotional—to recuperate from a significant trauma, even a chosen and positive one. Your body is directing a huge amount of energy toward rebuilding tissue. The lingering effects of anesthesia, powerful pain medications, and your body’s own inflammatory healing response can create a very real “brain fog” that makes concentration, memory, and complex thought feel impossible. This isn’t a personal failing; it is a physiological reality. Trying to push through it by forcing yourself to answer emails or study for an exam will not speed up your recovery; in fact, the stress it generates can actively impede it by flooding your system with cortisol, a hormone that interferes with the healing process.

Give yourself explicit and unapologetic permission to rest. For these few weeks, your only job is healing. Frame it as the most important and productive task you can possibly undertake. This is the time to lean on your support network and to practice the vulnerable art of asking for and accepting help. Be specific in your requests. Instead of saying “I might need some help,” try “Could you come over on Tuesday to do a load of laundry and make sure I have enough to drink?” Accepting this care isn’t a sign of weakness; it’s a vital part of a successful recovery strategy. Every hour you spend resting, every moment you choose quiet over activity, and every time you allow someone to care for you is a direct and powerful investment in your long-term health and the beautiful, congruent life you are building for yourself.

The Healing Timeline

Immediately after surgery, you’ll wake up in a recovery room where your medical team will be closely monitoring your vital signs, oxygen levels, and pain. It’s normal to wake up feeling groggy, disoriented, and cold. Expect significant swelling and bruising around the surgical site, which can look alarming but is a normal part of the body’s healing response. Your comfort is a top priority, and your pain will be actively managed, likely starting with medication through an IV and transitioning to oral pills as you become more alert.¹ To allow your new urethra to heal without any stress, a Foley catheter will be in place to drain your bladder. This might feel a bit strange, but it’s a crucial temporary measure. It’s typically removed after a couple of days, once your team is confident you can urinate on your own without difficulty.¹

The initial, most intensive recovery period, where you’ll be focusing almost entirely on rest, is typically about 4 to 6 weeks. It’s vital to remember that this is just an average; everyone heals at their own pace, influenced by factors like your age, overall health, and how closely you follow aftercare instructions.

During these first weeks, you will have regular follow-up appointments to monitor your healing, check your incisions, and ensure everything is progressing as it should. Your main job is to rest and follow your aftercare instructions to the letter. This will include keeping the area clean and dry, avoiding strenuous activity—absolutely no heavy lifting (nothing more than 5-10 pounds), no intense exercise, and no submerging the area in a bath, hot tub, or pool, as this can introduce bacteria and lead to infection. Short, gentle walks will be encouraged right away after surgery to promote circulation and prevent blood clots. Most patients can gradually and gently resume sexual activity in 2-3 months, once cleared by their surgeon and, most importantly, when they feel physically and emotionally ready.¹ It’s also crucial to remember that if your surgery was more extensive and included internal procedures like a hysterectomy and vaginectomy, the recovery period will naturally be longer as your body has much more internal healing to do.²

As with any major surgery, there are potential risks, and it’s important to be informed and vigilant. The general risks include bleeding, adverse reactions to anesthesia, blood clots (especially in the legs, known as DVT), and infection.¹⁰ Your team will give you specific signs to watch for, such as a fever over 101°F, increasing pain or redness that spreads from the incision site, or unusual, foul-smelling discharge. A specific long-term concern with nullification is the risk of urethral stricture, which is a narrowing of the urethral opening over time that can make urination difficult, causing a weak stream or a feeling of incomplete emptying.⁶ This is why the surgeon’s technique and experience are so critical. By following all of your surgeon’s post-operative instructions—such as taking prescribed antibiotics, keeping the area clean, and attending all follow-up visits—you can significantly reduce these risks and set yourself up for a smooth and successful healing journey.

Redefining Sexuality and Sensation

Nullification profoundly changes your body’s physical landscape, and this will naturally reshape your sexual experiences. It’s essential to frame this not as a loss, but as a liberation—a freeing of your intimate self from a cage it may have been in for years. For so many, intimacy before surgery is shadowed by the constant, draining presence of gender dysphoria. This isn’t just a fleeting thought; it’s a persistent cognitive load, a background static of wrongness that makes being truly present with a partner almost impossible. It’s the voice that makes you dissociate, turning you into an observer of your own life rather than a participant. It’s the anxiety about how your body looks, feels, or even sounds. The mental energy consumed by this internal conflict is immense. After surgery, that static is gone. The silence it leaves behind isn’t empty; it is a vast, peaceful, and open space that can finally be filled with genuine connection, presence, and pleasure.

This newfound peace invites you to completely redefine intimacy, freeing it from the narrow confines of traditional, often penetrative, acts that may never have felt right in the first place. You and your partners are given a blank page, an opportunity to co-author a new sexual story that is uniquely yours.⁸ This is about throwing out the old scripts—the ones often tied to binary gender roles and expectations—and starting fresh. It’s a journey of discovery built on open communication, curiosity, and joyful exploration. The entire body becomes a landscape of potential pleasure. This is the time to explore what feels good everywhere—the sensitivity of the inner thighs, the electricity of touch on the neck, the deep intimacy of full-body contact, the varying sensations of pressure, texture, and temperature. It’s an opportunity to learn your body’s new language of pleasure, a language that was always there but was previously drowned out by the noise of dysphoria.

For AMAB individuals who opt for a “hidden” clitoris, a powerful new center for orgasmic pleasure is created, one that is responsive to pressure and vibration through the skin.⁷ But that is just one possibility among many. The skin of the mons pubis and perineum can become newly sensitized and erogenous. For everyone, it’s a reminder that our largest and most powerful sex organ is our brain. By removing the primary source of dysphoria, you free up incredible neural bandwidth to simply feel. Personal stories from those who have undergone nullification often speak of an overwhelming sense of “rightness” and a profound gender euphoria that permeates every aspect of life, especially intimacy.¹¹ This feeling of being truly at home in your body for the first time is a powerful aphrodisiac. Before, you may have felt like a guest in your own skin, disconnected and performing a role. Now, you are the host, fully present and embodied. This fundamental shift can radically lower inhibitions, deepen emotional connection, and allow for a level of vulnerability with a partner that was previously unimaginable. You are no longer performing intimacy; you are embodying it. This newfound congruence is often the key that unlocks a more present, joyful, and deeply satisfying intimate life, one built on the incredible pleasure of finally being yourself.

Life as a Nullo: Navigating the Everyday

Living as a nullo is about more than just a physical change; it’s about navigating the world in a body that is finally in harmony with your mind. This newfound congruence can be a quiet but profound source of peace that touches every part of your daily life. After the intense recovery period, daily activities resume, but often with a new sense of comfort and ease that may have been absent for years. For AMAB individuals, urinating while seated becomes the new norm, a practical change that for many is also an affirming act that feels more aligned with their identity.¹ But the biggest changes are often subtle, felt in the private moments that make up a life. It’s the simple act of waking up and not having your first thought be a flash of discomfort with your own body. It’s showering without averting your eyes or feeling a sense of detachment. It’s catching a glimpse of your reflection in a window and feeling a click of rightness, not a jolt of wrongness. It’s about existing in your own skin without the constant, low-level hum of dysphoria—a draining background static you may not have even realized was consuming so much of your energy until it was finally gone.

This internal peace translates into a newfound freedom in the external world. Getting dressed, for instance, can transform from a daily chore or a source of anxiety into an opportunity for joyful self-expression. Before surgery, choosing clothes may have involved a complex and exhausting mental calculus: “Will this hide what I need it to hide? Will this fabric create a bulge? Will I feel exposed or fake in this?” The smooth groin area removes these physical and psychological barriers, opening up a vast world of clothing choices. Styles that may have been uncomfortable or dysphoria-inducing before—like leggings, swimwear, yoga pants, fitted trousers, or different cuts of underwear—can now be worn with confidence and comfort.¹² This freedom allows for a much more nuanced and authentic presentation. You can explore a traditionally masculine or feminine silhouette, or you can craft a uniquely androgynous look that feels true to you, one that relies on clean lines and a sleek form rather than hiding your body under baggy clothes. You become the driver with full creative control over how you present your shape to the world each day.¹³

Of course, living in a world that is still learning about gender diversity requires its own kind of thoughtful approach. Navigating social situations can be complex. Public understanding of non-binary and agender identities is growing, but it is far from universal. This can create challenging and stressful situations, from the dreaded public restroom dilemma to dealing with intrusive questions from strangers or even well-meaning acquaintances. You may find yourself doing a quick mental calculation before entering a public space: Which restroom is safer for me in this moment, based on how I’m presenting today? Is there a gender-neutral or family option available? Because of this, some individuals may choose to present in certain ways publicly as a of self-protection and energy conservation. It’s not about being inauthentic; it’s a strategic choice to opt for looser clothing or use specific facilities to avoid potential harassment and get through your day with less stress.⁶

Building and maintaining a strong support system hinges on clear, honest communication, which is itself a skill you will hone. With romantic partners, this means having ongoing, open conversations about boundaries, pleasure, and how you want to explore intimacy together. It’s a journey of mutual discovery. With friends and family, it might involve explaining your journey in a way they can understand, clarifying your pronouns, and letting them know how they can best support you—which often just means listening and continuing to love you. It is crucial to remember that you are in complete control of your own story. You decide who you tell, when you tell them, and how much detail you share. Your surgical history is intensely private information, and you are never obligated to disclose it to anyone you don’t want to. Choosing who to let into your inner world is a powerful act of self-care and boundary-setting. You might be an open book with your closest friend, share the basics with a new partner when you feel safe, and tell a curious coworker, “Thanks for your interest, but I don’t discuss my private medical history.” Each of these choices is valid, and each one is yours to make.

Finding and Building Your Support Network

You don’t have to walk this path alone. In fact, building a strong, multi-layered support network is one of the most critical parts of your surgical plan—as essential as choosing your surgeon or preparing your home for recovery. This isn’t just about having people to lean on; it’s about creating an ecosystem of care that can hold you through every stage of this profound transformation, from the anxious waiting period beforehand to the complex emotional and physical healing that follows.

Online communities can be an invaluable and immediate lifeline, especially when your real-world community lacks people with similar lived experiences. They are vibrant, 24/7 hubs of shared knowledge where you can connect with other trans and non-binary people who have either walked this exact path or are on it alongside you.² These spaces offer a unique kind of support you can’t get anywhere else. It’s where you can ask incredibly specific questions—about a particular surgeon’s technique, what a certain post-op sensation feels like, or how to deal with an unsupportive family member—and get answers from people who truly get it. It’s a place to see real recovery timelines and photos, which can demystify the process and provide a hopeful roadmap. This peer-to-peer connection combats the intense isolation that can come with a less common gender-affirming surgery. For moments of acute emotional distress, organizations like The Trevor Project¹⁵ and Trans Lifeline¹⁶ provide essential, free, and confidential crisis support. Critically, these services are run by and for the LGBTQ+ community, meaning the person on the other end of the line understands the unique pressures and fears you might be facing, offering a safe and affirming space to land when you need it most.

Learning to advocate for yourself within the medical system is another powerful and necessary of self-support. This means seeing yourself as an active and essential member of your own care team. It’s about doing your research, coming to consultations with a list of questions, and not being afraid to ask for clarification if you don’t understand a medical term or a post-op instruction. It’s about feeling empowered to say, “That doesn’t feel right for me,” or to seek a second opinion if a provider’s approach doesn’t align with your goals. When facing financial hurdles, which are an unfortunate reality for so many, know that there are community-based resources designed to help. Organizations like Point of Pride offer surgical scholarships,¹⁸ and platforms like FOLX Health can point you toward mutual aid funds and other financial resources.¹⁹ Applying for this kind of assistance is not a sign of failure; it is a courageous act of community care and self-advocacy.

Finally, and perhaps most importantly, consciously and intentionally nurture your personal support system. This is your chosen family—the friends, partners, and loved ones who see you, respect you, and celebrate you for exactly who you are. This requires proactive, honest conversations before surgery. Sit down with your primary caregiver(s) and talk frankly about what recovery will entail: the physical limitations, the emotional ups and downs, the need for help with everything from meals to medication reminders. This isn’t just about logistics; it’s about preparing them for the reality of caregiving so they can support you without burning out. Their understanding and steady presence will be your anchor in the occasionally turbulent waters of healing. Remember, your journey is valid, your needs are legitimate, and you are absolutely deserving of being surrounded by unwavering love and support every single step of the way.

Conclusion: One Path Among Many

The journey toward a body that feels like home is not a single road, but a vast landscape with countless paths. This chapter has offered a detailed look at one of those paths: nulloplasty. It is a powerful, valid, and deeply affirming choice for many who find their truth in neutrality—a radical act of self-definition that says, “My wholeness is not defined by what is present, but by a quiet and intentional peace.” This choice is a destination in its own right, not a halfway point between two other places.

It’s essential, however, to hold this knowledge as one piece of a much larger puzzle. Think of gender affirmation not as a linear checklist, but as the art of creating a mosaic of your own life. Each piece is a different choice: a name, a pronoun, a piece of clothing, a hormone, a surgery, a conversation with a loved one. For some, nulloplasty may be a large, central piece of their mosaic; for others, it may not be part of the picture at all. There is no finished image you are trying to copy. You are the artist, creating your masterpiece as you go, and only you can know when it feels complete and true.

Your journey is uniquely yours, and this cannot be overstated. The right path is whatever combination of steps—or lack thereof—leads you to a place of internal peace and authentic self-expression. There is no “right” way to be trans or non-binary, and there is certainly no required set of medical procedures you must have to be valid. Your affirmation may be entirely social and legal. It may involve hormones but no surgery. It may involve multiple surgeries. The only person whose validation truly matters is your own. The goal is never to meet an external standard, but to cultivate an internal sense of rightness.

As the following chapters explore other surgical pathways like vaginoplasty and phalloplasty, hold them with this same perspective. And if you’re here not for yourself, but as an ally—a parent, a partner, a friend, a therapist—seeking to better understand and support someone you love, we are so glad you are here, too. Your willingness to learn is a profound act of love in itself, and this information is just as much for you.

Perhaps after reading this chapter, you feel a deep resonance, a sense that nulloplasty is the right path for you. That is a powerful feeling. Even so, I encourage you to keep reading. Understanding the full spectrum of possibilities doesn’t invalidate your own feelings; it enriches them and gives you a broader context for your own choices. These chapters are not a menu to order from, but a continuation of this exploration—a look at the different tools available for different people with different goals. Read about them with curiosity, not pressure. See them as more possibilities in the vast landscape of what it means to be you. Whatever direction your journey takes, know that every single step you take in honoring your truth—whether that’s a major surgery or simply correcting someone on your pronouns—is a courageous and beautiful act of self-love. It is the work of coming home.

Chapter 7 Works Cited

1. Bookimed. (n.d.). Gender nullification surgery for non-binary individuals. https://us-uk.bookimed.com/article/nullification-surgery/

2. QueerDoc. (n.d.). Your smooth bottom line: Nulloplasty/Nullification/Nullectomy. https://queerdoc.com/nullectomy-nullification/

3. Wikipedia. (2025, October 5). Eunuch. https://en.wikipedia.org/wiki/Eunuch

4. AMA Journal of Ethics. (2023, June). Ethically navigating the evolution of gender affirmation surgery. https://journalofethics.ama-assn.org/article/ethically-navigating-evolution-gender-affirmation-surgery/2023-06

5. Post Street Surgery Center. (n.d.). Non-binary. https://www.poststreetsurgery.com/gender-affirmation/non-binary/

6. MacPhee, K. (n.d.). Non-binary genital reconstruction. https://keeleemacpheemd.com/non-binary-genital-reconstruction/

7. Davis Plastic Surgery. (n.d.). Genital/gender nullification. https://www.davisplasticsurgery.com/genitalgender-nullification.html

8. District Plastic Surgery. (n.d.). Gender nullification surgery. https://www.districtplasticsurgery.com/plastic-surgery-procedures-washington-dc/gender-affirmation/gender-nullification/

9. Crane Center for Transgender Surgery. (n.d.). Non-binary surgery. https://cranects.com/non-binary-surgery/

10. Oregon Health & Science University. (n.d.). Gender-affirming surgery. https://www.ohsu.edu/transgender-health/gender-affirming-surgery

11. Riggs, D. W., Pearce, R., Pfeffer, C. A., Hines, S., White, F., & Ruspini, E. (2025). “It feels like a lightness, and it feels… happy”: Qualitative exploration of transgender and non-binary adults’ understanding of gender euphoria, gender dysphoria, and sexual wellbeing. International Journal of Transgender Health.

12. ART Surgical. (n.d.). Gender nullification surgery. https://www.artsurgical.net/gender-nullification-surgery

13. FOLX Health. (n.d.). Gender-affirming clothing brands to support year-round. https://www.folxhealth.com/library/gender-affirming-clothing-brands-to-support-year-round

14. The Trevor Project. (n.d.). A guide to being an ally to transgender and nonbinary youth. https://www.thetrevorproject.org/resources/guide/a-guide-to-being-an-ally-to-transgender-and-nonbinary-youth/

15. Trans Lifeline. (n.d.). https://translifeline.org/

16. British Medical Association. (n.d.). Inclusive care of trans and non-binary patients. https://www.bma.org.uk/advice-and-support/equality-and-diversity-guidance/lgbtplus-equality-in-medicine/inclusive-care-of-trans-and-non-binary-patients

17. Point of Pride. (n.d.). Life-changing healthcare access for trans folks. https://www.pointofpride.org/

18. FOLX Health. (n.d.). Mutual aid funds. https://www.folxhealth.com/library/mutual-aid-funds

Every chef, whether professional or a passionate home cook, has a version of “The Pile.” It’s a sacred, chaotic testament to a life lived through food.

Mine is a teetering stack of culinary history, a physical archive of flavor. It’s the stained, flour-dusted pages of my grandmother’s favorite cookbooks, their spines cracked and pages softened from decades of love and use. It’s my personal collection of Moleskine kitchen notebooks, filled with frantic, sauce-splattered handwriting from years of professional service—a shorthand of temperatures, timings, and last-minute inspirations that only I can decipher. It’s a bulging manila folder of magazine clippings, a camera roll full of hastily snapped screenshots from obscure food blogs, and a scattered collection of digital notes-to-self about that one perfect vinaigrette I managed to reverse-engineer from a memorable trip to France.

This collection is, without exaggeration, my life’s work. It’s a treasure trove of flavor and memory, a map of my personal and professional evolution. It’s also a complete and utter mess. For years, this disorganization has been a source of low-grade, constant anxiety.

I’ve wrestled with an age-old problem that plagues every cook: How do I get all of this invaluable, disparate information into a single, instantly searchable, and genuinely useful place? My dream has always been to ask a simple question—”What was that braised short rib recipe from the 2008 notebook that used star anise?”—and get an immediate, actionable answer, not spend twenty minutes digging through a dusty box in the basement, pulling out notebook after notebook while the pot on the stove threatens to scorch.

Today, I finally started the project to solve this for good. My past attempts have been a graveyard of good intentions. I’ve tried manual transcription into spreadsheets, a soul-crushing task that never lasted more than a week. I’ve experimented with clunky scanning apps that produced blurry, unsearchable PDFs, which were somehow even less useful than the paper originals. I’m a long-time user of Paprika 3, and I truly adore its capabilities for meal planning, generating smart grocery lists, and instantly scaling portions. The problem was never the app itself, but the immense, tedious bottleneck of getting my analog life into it. Who has the time or the patience to manually create structured recipes out of every single handwritten card from grandma’s old tin vault? The friction was too great. But now, AI creates the pathway I’ve been missing, a bridge over that chasm of manual data entry.

The core of this new project is a custom script I’m developing, a little software “gem” I’ve affectionately named the “Recipe Structurizer.” Its sole purpose is to look at an image of a recipe—any recipe, in any format—and understand not just the words, but the underlying structure.

What makes it a game-changer is its flexibility and intelligence. It can process everything from a fully-articulated, professionally typed recipe in a spreadsheet to a vague, hand-written note on a stained napkin that just says “Chicken, lemon, thyme, high heat.” It doesn’t just pull out the ingredients and instructions that are explicitly listed; it can intelligently infer what’s missing, filling in the blanks like an experienced prep cook. If a recipe doesn’t mention the necessary pots, pans, and whisks, the gem analyzes the steps and generates a probable equipment list. If the instructions are just a single, cryptic sentence, it expands them into a clear, logical, step-by-step process that anyone could follow. It expertly turns chaotic, incomplete, and varied information into clean, usable, and consistently formatted structured data.

This is where the magic of the workflow truly begins, where friction gives way to flow.

The output from my “Recipe Structurizer” Gem isn’t just a jumble of text; it’s perfectly organized and tagged data. From there, my process is designed to be beautifully simple and automated. With a single click, I export the structured text directly into a new, neatly formatted Google Doc. Each recipe gets its own file, creating a uniform, personal digital cookbook as I steadily work through my pile. This growing library of recipes, stored securely in Google Drive, will become my go-to culinary memory, a digital extension of my own brain that is always available to Gemini. By simply using the @ connection to my Drive, I’m essentially giving my AI assistant a perfect, ever-expanding memory of my entire cooking journey. Plus, because the text is so clean and well-structured, getting it into Paprika becomes a trivial matter of copy and paste. Making this ancestral, handwritten knowledge machine-readable is more than half the battle.

The ultimate goal is to completely transform my daily interaction with my own culinary knowledge. Soon, instead of spending precious prep time digging through “The Pile,” frantically flipping through notebooks for a single ingredient substitution or a forgotten temperature, I’ll just ask. The query can be as simple as, “Show me that braised short rib recipe from the 2008 notebook,” and the full, structured recipe will appear on my screen instantly. But the real, world-changing power comes from the conversational depth this system enables. I’ll be able to scale an old, complex recipe for a dinner party on the fly—no more mental math or scribbling messy conversions on a scrap of paper. “Take Grandma’s lasagna that serves six and scale all ingredients for fifteen people,” and it will be done, accurately and instantly. Even more exciting is the boundless potential for creativity and innovation. I’ll be able to ask it to combine techniques from two completely different concepts, using it as a true digital sous-chef. “What would happen if I used the brine from my buttermilk fried chicken recipe on the pork chops from that old German cookbook? Generate a new recipe that merges the two.” It becomes more than just a static storage system; it becomes a dynamic, interactive tool for culinary exploration.

This is day one, and it’s a long work in progress, not a finished, polished archive. But for the first time, I have a clean, efficient, and even enjoyable workflow that can handle any type of recipe I throw at it. It’s the beginning of a real, lasting solution to a problem I’ve had my entire career. The pile is still there, sitting on my desk, but it no longer feels like an insurmountable mountain of work. It’s an archive-in-waiting, a queue of memories ready to be digitized. The real, living library is now being built, one recipe at a time.

Not bad for a new tool created from start to finish before noon without any code.

Subscribe now