The New York City Council held a hearing on AI safety on October 5. Whistleblowers warned of extinction-level risks from uncontrolled AI development.

Jacob Coxon, a former Anthropic researcher, said it is ‘more likely than not’ that humanity loses control of AI systems. He called the leading labs ‘extremely reckless’. Coxon quit Anthropic in September. He accused the company and OpenAI of ‘gambling with our lives’.

Daniel Kokotajlo, a former OpenAI researcher, said AIs are becoming superhuman at hacking. Alex Turner, a former Google DeepMind researcher, put the chance of AI takeover at roughly one in three. All three testified before the council’s Committee of the Whole.

Company representatives also testified. OpenAI, Anthropic, Google and Meta sent policy executives. They faced questions under oath from council members. Speaker Julie Menin convened the hearing at Coxon’s request.

The council is considering 10 bills. They include a kill switch for AI systems, whistleblower protections and deepfake restrictions on elected officials. Musk’s SpaceXAI did not attend despite a subpoena.