New York just put the AI labs under oath. Washington gave them a rebrand.

Today, something happened that Congress hasn't managed in two years of hearings-that-aren't-hearings: New York City's entire 51-member council put the AI industry's leaders under oath. Not a polite panel. Not a roundtable. Sworn testimony. On one side of the table sat the labs' safety and policy chiefs — Anthropic's head of its Frontier Red Team, OpenAI's policy development lead, plus AI policy executives from Meta and Google. Across from them sat the defectors: former Anthropic researcher Jacob Coxon, ex-DeepMind's Alex Turner, and ex-OpenAI researcher Daniel Kokotajlo, who now runs the AI Futures Project. Coxon didn't soften it for the cameras. He called the companies "extremely reckless," said the startup mantra of move-fast-break-things "does not work for the most powerful technology ever built," and testified that researchers no longer fully understand why AI systems do what they do — even as those systems approach the ability to improve themselves. Turner put the odds of an AI takeover at roughly one in three. Let that sit for a second. And the council didn't stop at talk. On the table: a bill requiring every AI system deployed in the city to pass third-party validation and carry a human kill switch, with $25,000 penalties per violation; payouts for whistleblowers who report dangerous practices; and a right for ordinary people to sue when bypassed safety controls cause harm. Elon Musk's SpaceXAI was served with an actual subpoena after it didn't even bother answering the invitation. Here's what gets me. Yesterday the federal government announced an AI task force and a new vocabulary word. Today a city council reached for the off switch. The most aggressive AI regulation in America isn't coming from Washington anymore — it's coming from City Hall. Sources: NY Post, "Anthropic whistleblower Jacob Coxon doubles down on AI warnings at NYC hearing: 'Extremely reckless'" (Oct 5, 2026) — https://nypost.com/2026/10/05/business/anthropic-whistleblower-jacob-coxon-to-testify-at-nyc-ai-hearing-on-kill-switches-regulations/ ; Bloomberg (via Moneyweb), "Ex-Anthropic researcher Coxon to appear at NYC hearing on AI" (Oct 5, 2026) — https://www.moneyweb.co.za/news/ai/ex-anthropic-researcher-coxon-to-appear-at-nyc-hearing-on-ai/ ; Reuters, "Former Anthropic researcher Coxon to testify at New York City AI hearing, Bloomberg News reports" (Oct 4, 2026) — https://www.reuters.com/business/former-anthropic-researcher-coxon-testify-new-york-city-ai-hearing-bloomberg-2026-10-04/ ; NY Post, "OpenAI, Google, Anthropic agree to attend NYC AI hearing after subpoena threats" (Sept 29, 2026) — https://nypost.com/2026/09/29/us-news/openai-google-anthropic-agree-to-attend-nyc-ai-hearing-after-subpoena-threats/