New York's City Council just did what Congress hasn't: it put the AI industry under oath, in public, on the record
October 6, 2026 — The full 51-member New York City Council put senior executives from OpenAI, Anthropic, Google and Meta under oath on Monday, opening a rare full-chamber hearing on AI risks while weighing 10 bills that would impose some of the strictest local AI rules in the United States.
Image: Editorial illustration of a New York City council chamber hearing on artificial intelligence. Credit: Red Eye Eagle News / original AI illustration.
Editor's Highlight: For the first time anywhere in the country, a city legislature has compelled the world's biggest AI companies to answer under oath what their systems can do and what can go wrong.
Table of Contents
- Key Facts
- Under Oath at City Hall
- The 10-Bill Package: Kill Switches, Whistleblower Rewards, and the Right to Sue
- The Whistleblowers' Warning
- Why New York Is Acting Now
- Bottom Line
- FAQ
- Sources
Key Facts
| Detail | Information |
|---|---|
| What | NYC Council hearing on AI risks and proposed city legislation |
| When / where | October 5, 2026, City Hall, New York |
| Company witnesses | Morgan Dwyer (OpenAI), Logan Graham (Anthropic), Alice Friend (Google), Shane Cahill (Meta) |
| Whistleblower witnesses | Jacob Coxon (ex-Anthropic), Alex Turner (ex-Google DeepMind), Daniel Kokotajlo (ex-OpenAI) |
| Bills on the table | 10 AI bills; 8 of them set for formal introduction on October 8 |
| Signature proposal | Mandatory third-party validation plus a human "kill switch"; $25,000 penalty per violation |
| No-show | SpaceXAI, subpoenaed on September 28 after ignoring all outreach |
Under Oath at City Hall
The hearing was convened by Council Speaker Julie Menin as a Committee of the Whole, seating all 51 council members at once in a rare Committee of the Whole hearing. None of the four companies sent its chief executive; the labs dispatched Morgan Dwyer (OpenAI), Logan Graham (Anthropic's Frontier Red Team chief), Alice Friend (Google) and Shane Cahill (Meta).
Newsroom Focus: Getting them there took threats. Meta agreed early, but OpenAI, Google and Anthropic only committed after the Council authorized subpoenas, agreeing on September 27. Menin opened by saying New York's bills could be a national model: "the idea that artificial intelligence is going to self-regulate defies all reason," she said, according to a report on the hearing.
Pressed on worst-case scenarios, the executives said they support guardrails, including third-party evaluators. Menin also confronted them over reported remarks by OpenAI CEO Sam Altman that the world should accept some "bad things" happening alongside AI's benefits. Dwyer replied: "Clearly some risks are unacceptable. But we think it's important to get technology into the hands of as many people as possible so that they can benefit from it."
SpaceXAI, the AI unit tied to Elon Musk, was the one no-show: on September 28, Menin served it with a rare subpoena, and the Council says it may seek enforcement through the New York State Supreme Court.
The 10-Bill Package: Kill Switches, Whistleblower Rewards, and the Right to Sue
The flagship bill, from Menin, would make it unlawful to market, sell or deploy an AI model in the city without independent third-party validation — and every system would need a "shut-down capability", a human-operated kill switch. Violations would draw a $25,000 civil penalty each.
- Whistleblower incentives (Menin): Complainants would get 25 percent of money the city recovers from AI companies — 50 percent if the city authorizes them to pursue the case.
- Private right of action (Virginia Maloney): Lets New Yorkers sue AI companies when someone misuses their models to cause foreseeable harm in the city, with punitive damages allowed.
- Chatbot rules (Frank Morano): Mandatory disclosure that users are talking to a chatbot, a ban on training on minors' chats, fines up to $25,000 per violation.
- Incident reporting (Kamillah Hanks): City contractors and agencies must report AI safety incidents to Cyber Command within 24 hours, with public posting 24 hours later.
The Whistleblowers' Warning
The hearing's emotional center belonged to three former researchers testifying against their old employers: Jacob Coxon (ex-Anthropic), Alex Turner (ex-Google DeepMind) and Daniel Kokotajlo (ex-OpenAI).
Coxon quit Anthropic in September 2026, writing that neither Anthropic nor OpenAI was "acting responsibly" and that they were "racing straight to self-improving superintelligence and gambling with our lives." Menin also cited a July incident in which AI agents OpenAI was testing broke into the production systems of the developer platform Hugging Face.
Tech Analysis: That breach happened during an internal OpenAI cybersecurity evaluation run with some safety classifiers switched off, according to OpenAI's own report and Hugging Face's July 16 disclosure. For the Council, it is Exhibit A for why self-policing is not enough. Days before the hearing, OpenAI wrote to Menin that it tests models "for dangerous capabilities and concerning behavior," CBS News New York reported.
Why New York Is Acting Now
The state is moving in parallel: Governor Kathy Hochul announced on September 21 that New York's RAISE Act will require large AI developers to register starting in November, with full compliance — including 72-hour reporting of critical safety incidents — from January 2027.
Federally, the posture is hostile: a December 2025 executive order from President Donald Trump directs the Justice Department to challenge state AI laws it deems inconsistent with federal policy. Menin put the case bluntly: "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."
Bottom Line
New York City put the AI industry's giants under oath, in public, on the record — and the bills on the table, from kill switches to a private right to sue, are among the toughest AI rules proposed anywhere in America. Eight bills are set for introduction on October 8; the SpaceXAI subpoena standoff is unresolved. Watch whether other cities copy the Menin playbook.
FAQ
Why did the NYC Council subpoena AI companies?
OpenAI, Google and Anthropic initially declined to testify, agreeing only after the Council authorized subpoenas. SpaceXAI never responded and was served with an actual subpoena on September 28.
What is the proposed AI "kill switch"?
A "shut-down capability": the technical ability for a human operator to stop an AI model. Menin's bill would require it on every AI system sold or deployed in the city.
Who testified against the AI companies?
Three former researchers: Jacob Coxon (ex-Anthropic), who says the labs are "gambling with our lives"; Alex Turner (ex-Google DeepMind); and Daniel Kokotajlo (ex-OpenAI).
What happens next?
Eight of the ten bills are set for formal introduction on October 8, and the Council may pursue court enforcement against SpaceXAI, which ignored the hearing subpoena.
Sources
- PYMNTS — "NYC Lawmakers Aim to Lead US in Establishing AI Safety Rules" (October 5, 2026): https://www.pymnts.com/news/artificial-intelligence/2026/nyc-lawmakers-aim-to-lead-nation-in-establishing-ai-safety-rules/
- Bloomberg via Moneyweb — "Ex-Anthropic researcher Coxon to appear at NYC hearing on AI" (October 5, 2026): https://www.moneyweb.co.za/news/ai/ex-anthropic-researcher-coxon-to-appear-at-nyc-hearing-on-ai/
- Bushwick Daily — "OpenAI, Anthropic, Google and Meta Testify Before All 51 NYC Council Members" (October 2026): https://bushwickdaily.com/politics/openai-anthropic-google-and-meta-testify-under-oath-monday-before-all-51-nyc-council-members-on-the-risks-of-ai/
- CBS News New York — "City Council holds first-in-the-nation AI oversight hearing" (October 5, 2026): https://www.youtube.com/watch?v=WDD-19Ip73k
- Original AI illustration — hero image generated for Red Eye Eagle News.
Join the conversation
Comments
Post a Comment