NYC AI Hearing: Ex-Anthropic Researcher to Testify

Former Anthropic researcher Jacob Coxon is set to testify on Monday, October 5, 2026, at the NYC AI Hearing, where city lawmakers are weighing a sweeping package of artificial intelligence safeguards. He will appear at the request of Council Speaker Julie Menin, as reported by Bloomberg News on Sunday, October 4, 2026, in a story republished in a Reuters wire account.

Coxon resigned from Anthropic in September 2026 with a stark public warning, writing that the people building advanced AI earnestly believe the technology "could kill us all by the end of the decade," according to the Reuters account of the report. He accused his former employer and OpenAI of "gambling with our lives," as reported by the wire service. Reuters added that it could not immediately verify the Bloomberg report.

The NYC AI Hearing would give lawmakers direct input on a legislative package unveiled on September 25, 2026, that includes cash rewards for whistleblowers, a new right for residents to sue AI developers over foreseeable harms, mandatory third-party safety audits, and human-controlled kill switches on covered systems, according to the council's own announcement of the bills. Read more on the AI News desk.

How the NYC AI Hearing Took Shape

The road to the NYC AI Hearing began on September 9, 2026, when Coxon resigned and told the Wall Street Journal that AI systems could spiral out of control, according to reporting republished by AI Weekly. Days later, Anthropic's alignment science lead Evan Hubinger posted that he personally assigned a greater-than-ten-percent chance to AI causing human extinction within the decade, as reported by AI Weekly. Anthropic CEO Dario Amodei responded with an essay calling for a slowdown, not a halt, in frontier development, and nearly 1,400 AI-sector employees signed an open letter making a similar request, according to the same report.

Then the council got involved. Between September 15 and 17, 2026, Menin sent letters to the chief executives of five companies — Anthropic, OpenAI, Google, Meta and Elon Musk's xAI — requesting testimony under oath and setting a response deadline of September 25, according to reporting by Tech Times. Meta was the first to confirm it would send a senior executive; the others faced formal subpoena orders after the deadline passed, with the council reserving the right to enforce them through the courts, as reported by Tech Times. "Given the high stakes, these firms owe it to the public to come before the Council, answer our questions, and provide input on our proposed legislation under oath," Menin said in a council press release quoted by Tech Times.

What the NYC AI Hearing Could Change

The NYC AI Hearing centers on a package the council describes as one of the most ambitious local AI regulatory frameworks in the country. The broadest measure would bar any business from selling or deploying an AI system in the city without independent third-party validation, with reviewers assessing data quality, bias, decision outputs, privacy and security under standards set by the city's cyber command, according to the council's bill summaries republished by Harlem World Magazine. Every covered system would also need a human-operated kill switch, with penalties reaching twenty-five thousand dollars per non-compliant instance, as reported by the same source.

A second bill would make New York City the first American jurisdiction to pay whistleblowers a share of fines recovered from AI companies that break the law, while a companion measure would extend whistleblower protections to city employees and contractors who flag AI safety threats, according to the council announcement. A third proposal would let individuals harmed by AI tools sue the responsible companies directly when the harm was foreseeable, reasonable safeguards were missing, and a third party exploited the failure through so-called jailbreaking — a provision legal experts say could become the most consequential part of the package, as reported by Tech Times. A fourth bill would require city contractors to report AI safety incidents within twenty-four hours.

If the NYC AI Hearing goes ahead as reported, the session would put a city of more than eight million residents in direct conversation with the people building the most powerful models in the world. The hearing's real significance is jurisdictional: while Congress has passed no major AI safety law, cities and states are writing their own rules, as GenZ NewZ reported on the governors' bipartisan AI coalition building shared standards state by state. A whistleblower testifying under oath before all fifty-one council members would give that local movement its most dramatic witness yet.

Why the Numbers Behind the NYC AI Hearing Matter

The testimony would land at a moment when the industry's own safety researchers are publicly assigning large probabilities to catastrophic outcomes. Hubinger's greater-than-ten-percent estimate — posted in response to Coxon's resignation — is one of the most explicit existential-risk statements ever made on the record by a senior lab researcher, according to AI Weekly's reporting on the exchange. But the number has vocal critics: RAND science-and-technology policy researcher Michael Vermeer told Nature that such predictions rest on so many untestable claims the debate resembles faith more than science, as reported by AI Weekly, and Heidy Khlaaf of the AI Now Institute calls the extinction framing fear-mongering while pointing to the reliability of AI systems in critical environments instead.

Whatever happens at the NYC AI Hearing, the session signals how seriously lawmakers now take the gap between what AI labs say privately and what they do publicly. Anthropic has confirmed neither the testimony nor any comment on Coxon's claims in the wire account, and OpenAI has not responded to the allegations either. Watch the hearing for whether the council moves the bill package forward — and whether other whistleblowers answer Menin's invitation to testify.