The New York City Council holds a hearing on AI risks on Monday, October 5. The Council says Anthropic, OpenAI, Google and Meta agreed to testify under oath. It subpoenaed Elon Musk’s SpaceXAI on September 28 after the company did not reply.
New to this? Read it in simple words
- New York City lawmakers question big AI companies in public on Monday.
- Anthropic, OpenAI, Google and Meta agreed to answer under oath.
- The city ordered SpaceXAI to take part after it did not reply.
- One bill would require a “kill switch” for AI systems used in the city.
- Subpoena
- A legal order to appear or hand over information.
- Under oath
- Promising to tell the truth, with penalties for lying.
- Kill switch
- A way for a person to shut an AI system down.
Who is testifying
Anthropic, OpenAI, Google and Meta agreed to testify publicly under oath, the Council said on September 28.
The Council subpoenaed SpaceXAI the same day because it had not answered the Council’s questions. The Council said it could ask a state court to enforce the subpoena.
Former Anthropic researcher Jacob Coxon will also testify, at Menin’s request, Bloomberg reported on Sunday. Reuters relayed the report but could not verify it.
Coxon quit Anthropic last month, warning that people building AI believe it “could kill us all by the end of the decade”, Reuters reports.
From the New York City Council’s releases of September 25 and 28, and Bloomberg’s report as relayed by Reuters.
The bills on the table
Menin’s main bill would require a “third party validation” before an AI system is sold or used in the city. It also requires a human override that can shut the system down.
Breaking it would cost $25,000 per instance, for both the business and the checker. A second Menin bill would let whistleblowers receive part of the fines.
Other bills would let people harmed by AI sue, make city contractors report AI safety incidents within 24 hours and require an emergency plan for AI incidents.
What to expect
The hearing is a Committee of the Whole, so all 51 Council members can take part. It is set for 11 a.m. in New York, according to the civic group BetaNYC.
“Leading experts and AI executives themselves are warning about the potentially grave risks of rapidly advancing AI,” Menin said in the September 28 release.
A hearing is not a vote. The bills still have to pass the Council, and a city-wide rule for every AI system may clash with state or federal law.
Sources
Every fact in this story comes from the sources below. Open them to check our work.
- 1Primary source · September 28, 2026Anthropic, OpenAI, Google, and Meta to Publicly Testify Under Oath New York City Council
- 2Primary source · September 25, 2026Speaker Menin unveils AI legislative package New York City Council
- 3Research · October 4, 2026Former Anthropic researcher Coxon to testify at New York City AI hearing, Bloomberg News reports Reuters
- 4
A preview written before the hearing: the companies, subpoena and bills come from the Council’s own releases, and Coxon’s testimony rests on a single Bloomberg report. We will update the story after the hearing.
New York City’s hearing on ten proposed AI bills is a political step, and no bill has been passed yet. How we rate the mood