The news
Lina Khan, who led the Federal Trade Commission until early 2025, has called on federal prosecutors to treat serious violations by AI companies the same way earlier regulators handled misconduct at large corporations. She pointed to a 1934 legal precedent as a model for holding individual executives accountable rather than stopping at corporate fines.
The remarks were reported on The Register and quickly reached the front page of Hacker News, where they drew 116 points and 55 comments within hours.
Context
Khan’s position rests on the claim that current federal law already contains tools to charge executives personally when their firms engage in fraud, antitrust violations, or other prohibited conduct. The 1934 precedent she cited established that corporate officers can face criminal liability for directing or knowingly permitting illegal acts.
Prior enforcement in technology and finance has largely relied on civil penalties and consent decrees. Khan argues that approach has not produced sufficient deterrence in fast-moving fields such as artificial intelligence. The shift she advocates would move cases from administrative settlements toward individual indictments when evidence shows knowing participation in violations.
This view aligns with patterns seen in earlier regulated industries. Banking and securities cases from past decades sometimes resulted in prison terms for officers who approved false statements or concealed risks. Khan presents the AI sector as no different in principle, even though the technology itself is new.
Details
The Register article notes that Khan believes “there are plenty of laws on the books to hold companies, and potentially their execs, accountable.” She did not name specific AI incidents or name current CEOs as targets. Instead she framed the issue as one of consistent application of existing statutes.
No new legislation is required under her reading. The 1934 precedent, whatever its precise origin, is presented as proof that criminal sanctions against individuals have long been available when corporate misconduct reaches a certain threshold. Prosecutors would still need to meet the usual standards of proof, including evidence of intent or willful blindness.
The article does not detail the exact statute or case from 1934. It focuses instead on Khan’s broader point that the Department of Justice already possesses the authority and has used it in other sectors. Khan’s comments appear to respond to ongoing debates about how to police claims made by AI developers around safety testing, data use, and market competition.
Reactions / counterpoints
The Hacker News discussion thread focused on whether such enforcement would chill innovation or simply raise the cost of cutting regulatory corners. Commenters did not surface additional public statements from Khan beyond the Register reporting. Several threads examined the practical hurdles prosecutors face when trying to prove that an executive knew a model’s claims were false or that a data practice violated existing rules.
No counter-statement from current FTC leadership or the Department of Justice appears in the coverage. The piece records Khan’s position without recording immediate replies from industry groups or sitting officials. The absence of named targets in her remarks leaves open the question of which conduct would cross the line from civil to criminal in practice.
Why it matters
For engineers and founders building AI products, the signal is that personal exposure may increase if a company’s claims about safety, data handling, or competitive conduct later prove false. Civil fines have historically been absorbed as a cost of doing business; criminal charges would change that calculation for the people who sign off on product decisions.
The argument Khan advances treats AI development as subject to the same accountability standards applied to earlier waves of technology and finance. It does not require proving that large language models pose unique physical risks. It requires only that prosecutors apply existing fraud, antitrust, or consumer-protection statutes when evidence shows executives directed or concealed violations.
Whether this approach gains traction depends on the first cases the Department of Justice chooses to bring. If prosecutors begin naming individuals in indictments, defense costs and insurance premiums for AI executives would likely rise. If they continue to settle at the corporate level, the practical effect of Khan’s remarks would remain limited to public debate.
The precedent she invoked has sat on the books for decades. Its use in the AI context would test whether enforcement agencies view software companies as subject to the same individual-liability rules that have applied to banks, drug makers, and industrial firms. That test will occur through specific charging decisions rather than through new statutes.
---
Sources:
{
"sources": [
{
"publisher": "Hacker News (front page)",
"title": "Ex-FTC boss Khan: break out the handcuffs for AI CEOs, citing 1934 precedent",
"url": "https://www.theregister.com/ai-and-ml/2026/09/14/ex-ftc-boss-khan-urges-uncle-sam-to-break-out-the-handcuffs-for-ai-ceos-citing-1934-precedent/5296325",
"published_at": "2026-09-15T00:40:14.000Z",
"summary": "Article URL: https://www.theregister.com/ai-and-ml/2026/09/14/ex-ftc-boss-khan-urges-uncle-sam-to-break-out-the-handcuffs-for-ai-ceos-citing-1934-precedent/5296325 Comments URL: https://news.ycombinator.com/item?id=49706223 Points: 116 # Comments: 55"
},
{
"publisher": "The Register",
"title": "Ex-FTC boss Khan urges Uncle Sam to break out the handcuffs for AI CEOs, citing 1934 precedent",
"url": "https://www.theregister.com/ai-and-ml/2026/09/14/ex-ftc-boss-khan-urges-uncle-sam-to-break-out-the-handcuffs-for-ai-ceos-citing-1934-precedent/5296325",
"published_at": "2026-09-14T16:59:24.000Z",
"summary": "There are plenty of laws on the books to hold companies, and potentially their execs, accountable"
}
]
}
No comments yet