Alex Shieh's anti-fraud AI startup turns 'snitching as a service' into a business model

Summary

Alex Shieh has launched an anti-fraud startup built around a simple premise: use AI and investigative journalism to blow the whistle on corporate fraud against the government, then collect a cut of whatever gets recovered. The company raised a $5 million pre-seed and seed round from Abstract Ventures, Router Ventures, and Dune Ventures, announced the same week as the interview.

The business model is contingency-based. Shieh calls it "snitching as a service" — the company earns nothing until a whistleblower case results in a government recovery, at which point federal whistleblower programs pay out a percentage of the recovered funds. There are no subscription fees, no software licenses.

The market

The Government Accountability Office estimates roughly $500 billion a year flows to fraud against U.S. government programs. Shieh's co-founder, Sah Sharda, authored The College Cartel, documenting how Ivy League universities allegedly price-fixed financial aid, a case that resulted in hundreds of millions of dollars in settlements. That's the kind of case the company was built to pursue systematically.

The immediate focus is big pharma. Co-founder David Barklay previously worked at the FTC under Lina Khan, where he was involved in efforts to prevent proprietary drug companies from blocking generic inhalers from entering the market. Healthcare is roughly 20% of U.S. GDP, and Shieh sees it as the first vertical before expanding into education and defense.

How it works

The operational model is closer to Palantir's forward-deployed engineer model than a conventional SaaS business — Shieh worked at Palantir before founding the company. Teams of journalists and AI engineers work cases together, with LLMs used to sift through the unstructured text in government filings and contracts. The goal is to automate more of that pipeline over time, but the current setup is human-led.

The journalism angle

Shieh frames this partly as a business model fix for investigative journalism. Ad-supported newspapers can no longer sustain the cost of long-form investigations, and most corporate fraud — diffuse, paperwork-heavy, with no single identifiable victim — doesn't translate into the kind of narrative that generates a book deal or a film. The Theranos story worked because Elizabeth Holmes was charismatic and the product was viscerally relatable. Pension fund skimming isn't. Shieh argues that tying journalist compensation to government recoveries solves both the funding problem and the incentive problem.

Shieh also has personal history here. While at Brown, he launched an investigation into the university's finances and testified before Congress about them.