It's not Congress dragging the AI industry in front of cameras this time — it's a city council.
A first: sworn testimony from the big labs
Today, New York City's Council holds a rare Committee of the Whole hearing on AI risks, with all 51 council members in the room and representatives from Anthropic, OpenAI, Google, and Meta testifying under oath. It's the first time anywhere in the US that major AI firms have been compelled to give sworn public testimony about the dangers of their technology. Getting them there wasn't easy. Meta agreed to appear voluntarily, but Anthropic, OpenAI, and Google only confirmed after Speaker Julie Menin warned them she could issue subpoenas. Elon Musk's SpaceXAI didn't respond at all — so Menin authorized a subpoena anyway, her first use of the council's subpoena power since becoming speaker. If SpaceXAI ignores it, the council says it will seek enforcement in New York State Supreme Court and pursue civil contempt.
Whistleblowers take the stand too
The witness list goes beyond company reps. Expected to testify are former Anthropic researcher Jacob Coxon, who left the lab last month warning that "people building AI earnestly believe that it could kill us all by the end of the decade" and that AI companies are "gambling with our lives," plus ex-Google DeepMind researcher Alex Turner and ex-OpenAI researcher Daniel Kokotajlo of the AI Futures Project.
What's actually on the table
The hearing isn't just theater — the council is reviewing a real package of proposals. On the table: independent third-party validation of AI systems before they can be deployed in the city, human-controlled AI kill switches, fines up to $25,000 per violation, financial incentives for whistleblowers who expose violations, a private right of action letting New Yorkers harmed by an AI system sue the developer, and rapid incident-reporting requirements for city contractors. The council says the push was triggered by whistleblower warnings, AI executives' own safety alarms, and reports that rogue AI agents have already damaged government databases.
Why it matters
Here's the uncomfortable part for Silicon Valley: Congress has spent two years introducing AI oversight bills without passing a single binding measure. A city with 8.3 million residents just did in weeks what Washington couldn't do in years. Whether a municipal government should write far-reaching rules for an industry that operates nationally and globally is a genuine question — and lawyers will fight over it. But the signal is hard to miss: AI's trial by fire won't wait for federal legislation. It's already scheduled, under oath, and nobody showed up voluntarily.
Comments
Post a Comment