NYC Council subpoenas SpaceXAI; Anthropic, OpenAI, Google, Meta to testify
New York City Council Speaker Julie Menin said Monday the Council secured under-oath testimony from Anthropic, OpenAI, Google, and Meta for an October 5 Committee of the Whole hearing on AI risks, and issued a rare subpoena to Elon Musk’s SpaceXAI after the company did not agree to appear.
A major American city is putting the main AI labs under oath in public next week, after rogue agents and safety warnings made the news. SpaceXAI is the holdout: it did not agree to appear, and the Council subpoenaed it.
On Monday, 28 September 2026, New York City Council Speaker Julie Menin announced that the Council had secured the first public testimony under oath from Anthropic, OpenAI, Google, and Meta. The subject is the potential dangers of AI, and what city laws might do about them. Under oath means the testimony is sworn. Those lines are the Council’s, in a press release dated the same day.
The testimony is set for Monday, 5 October 2026, at a Committee of the Whole hearing. That is a hearing of the full Council, not a smaller committee. The Council said all 51 members will sit, and it called the format rare. The release says the hearing follows weeks of warnings from whistleblowers and AI executives, and reports that rogue AI agents have already damaged government databases. An agent is software that can take actions, not only answer a question. Rogue, here, means it acted outside the limits people set for it.
Menin said: “Leading experts and AI executives themselves are warning about the potentially grave risks of rapidly advancing AI, and governments have a responsibility to act.” She said that is what the Council is doing. She said that, given the stakes, these firms owe the public an appearance, answers, and input on proposed legislation, under oath. That quotation is hers, in the Council release.
Who agreed, as the Council tells it. Between 15 and 17 September, Menin sent letters to five companies asking them to testify voluntarily. Meta then confirmed that a member of its senior leadership would appear. The deadline for an answer was Friday, 25 September. By that Friday, Google and Anthropic had declined. Aside from Meta, the companies that later said yes did so only after the Council’s explicit threat of a subpoena. A subpoena is a legal order to show up and testify. SpaceXAI, Elon Musk’s AI company, had still not responded. Menin issued a subpoena earlier Monday compelling the company to appear.
The rest of that clock, still the Council’s. On Friday, after Google and Anthropic declined, Menin authorized subpoenas to the four companies that had not yet agreed, to be issued as of 9:00 a.m. Monday. Through counsel, she then emailed those four a warning that a subpoena would come Monday. On Sunday, OpenAI and Google agreed to appear. Anthropic confirmed late Sunday night, hours before a subpoena was due to be served. SpaceXAI is the company that still had not agreed, and the one that received Monday’s subpoena.
CNBC reported the same Monday that the subpoena requires Musk, or another SpaceXAI representative, to testify. CNBC said a letter signed by Menin and a Council attorney, Nwamaka Ejebe, describes an investigation into whether fast-emerging risks to public safety, cybersecurity, economic stability, privacy, consumers, and businesses warrant immediate legislative action to protect New Yorkers. That description is CNBC’s account of the correspondence. The Council’s press release says the subpoena compels SpaceXAI to appear. It does not print the letter.
The hearing will also take testimony from national experts in AI safety, technology, and consumer protection. Members will review legislative proposals. A proposal is a bill that has been offered. It is not a law that has passed. The Council names a whistleblower incentive program, which would pay people who report problems from inside a company, and calls that program the first of its kind in the nation. It names a private right of action for New Yorkers harmed by AI agents, which would let a person sue over that harm instead of leaving enforcement only to the city. It names independent third-party validation, an outside checker rather than the company grading itself. The release says these sit among other bills. It does not say any company has admitted wrongdoing, and it does not say the October hearing has already produced a finding.
If SpaceXAI does not comply, the Council said it may ask a court to enforce the subpoena in New York State Supreme Court. That is New York’s state trial court, not the U.S. Supreme Court. The release points to Section 29 of the New York City Charter, which lets the Council investigate the city’s affairs and require people to attend and testify under oath, and to Council Rule 7.150, which lets the Council issue subpoenas for that work.
In plain terms, New York’s city legislature is putting the major AI labs under oath next week. When SpaceXAI would not agree to show up, the Council used a subpoena. The hearing is still ahead.
The photograph is an official picture of Speaker Julie Menin at the podium in City Hall chambers, from the Office of the Speaker. It shows Menin, the official who announced the hearing and issued the subpoena. It is not a picture of the subpoena, and it is not a picture of the October hearing.