Case files — Fortune audit · Case file 001 · 7 min read

They became billionaires at 22. Here's the machine that did it.

Fortunes Audited · Episode published Sep 24, 2026 · Built from public records

The thesis

Mercor's product isn't software. It's people: the knowledge inside the heads of doctors, lawyers and PhDs, packaged and sold to the companies building AI. A real fortune, with real exposures.

QualifiedFortunes Audited

22Age of the founders when they became billionaires, October 2025

$10BValuation of the October 2025 round, ~22% owned by each founder

$2BAnnualized revenue by June 2026, up from $100M in March 2025

60–70%Share of billings that passes through to contractors

The debate team

Brendan Foody, Adarsh Hiremath and Surya Midha were debate teammates at Bellarmine College Prep in the Bay Area. All three left college on Thiel Fellowships and, in 2023, founded Mercor: an AI tool that interviewed software engineers in India and matched them with American companies. Useful, but nobody becomes a billionaire at 22 running a recruiting marketplace.

The pivot to AI

By 2023 the AI labs had read most of the public internet. To get better at medicine, law or finance, they needed experts showing them, step by step, how professionals think. Mercor already had the machinery to find, test and hire skilled people fast, so it pointed that machine at a new customer: doctors, lawyers, bankers and PhDs for AI labs.

How the machine makes money

An AI lab pays Mercor an hourly rate for each expert. Mercor pays the expert less and keeps the difference: a staffing marketplace running at software speed. According to Tech Funding News, contractors take home roughly 60 to 70% of what's billed, so Mercor keeps roughly 30 to 40 cents of every dollar. Forbes reported annualized revenue of $100 million in March 2025, $500 million by September, $1 billion in February 2026 and $2 billion by June 2026.

The fortune is a valuation, not a bank balance. In October 2025, Mercor raised $350 million at a $10 billion valuation; Forbes estimated each founder owned about 22%, or roughly $2.2 billion each. In July 2026, Forbes reported talks at a $20 billion valuation.

The audit: five exposures

The headline number is gross, not net
If contractors keep 60 to 70%, Mercor's own net revenue is closer to $600 to $800 million. A $20 billion valuation on that is roughly 25 to 33 times what the company actually keeps.
A handful of customers
Named customers include OpenAI, Anthropic and Meta: a strength, and a concentration risk. After a data breach in March 2026, Tech Funding News reported that Meta paused all work with Mercor indefinitely.
Internal controls
Forbes reported an early hire who allegedly sent his own relatives hundreds of thousands of dollars in unauthorized bonuses (Mercor says it recovered the money), and suspected North Korean operatives who got past identity checks. For a company that sells verified expertise, verification is the product.
Data security
In late March 2026, Mercor was caught in a supply-chain attack through LiteLLM, an open-source package; public reporting says up to four terabytes of internal data and contractor information were exposed, and contractors filed class actions.
People and lawsuits
Rival Scale AI sued in September 2025, alleging a former employee took confidential documents; Mercor denies using them. In November 2025, Forbes reported some contractors were offered $16 an hour, down from $21. A marketplace has two sides; if experts stop trusting it, supply dries up.

The verdict

What's solid

The demand is real, the customers are the biggest AI labs in the world, the growth is among the fastest ever reported, and the founders spotted the human-expertise bottleneck early.

What's exposed

Headline revenue is mostly pass-through, a few customers carry the business, verification controls have failed more than once, and AI labs may one day need less human training data.

Qualified

A real fortune, measured in valuation, riding on a few customers and resting on controls that haven't yet matched its speed. Fix verification and widen the customer base, and it could be upgraded to Clean.

Chapters

Red flags in this case

Gross vs. net revenue · Annualized run-rate · Paper wealth · Customer concentration · Weak internal controls · Third-party risk

Sources

The audit opinion is the channel's opinion based on public reporting. Not financial advice.