On October 5, 2026, executives from Anthropic, OpenAI, Google and Meta testified under oath at a New York City Council hearing on artificial intelligence risks and possible safeguards. The council convened all 51 members for the hearing, alongside AI safety and consumer-protection experts.
The company representatives were Logan Graham, Anthropic’s head of Frontier Red Team; Morgan Dwyer, OpenAI’s head of policy development and operations; Alice Friend, Google’s director of AI and emerging tech policy; and Shane Cahill, Meta’s AI policy director for legislation. Former Anthropic researcher Jacob Coxon warned that AI companies were taking an “extremely reckless” approach. Former Google DeepMind researcher Alex Turner argued that misaligned AI could become a danger greater than foreign competitors. Those were warnings from the speakers, not findings established by the hearing.
Council Speaker Julie Menin pressed the company representatives on whether they could quantify the chance of a catastrophic event and on who should be legally responsible if AI caused serious harm. She said their answers on catastrophic risk were “troubling at best.” Google’s Friend called for a comprehensive federal framework, while Menin criticized the federal government’s light-touch approach and said AI companies could not be expected to regulate themselves.
The council is considering new safeguards
The hearing followed weeks of outreach by the council. According to the council’s September 28 announcement, Meta agreed to attend without a subpoena threat, while Anthropic, OpenAI and Google confirmed participation after the council warned it could subpoena them. The council had issued a subpoena to Elon Musk’s SpaceXAI, but no representative appeared. Menin said the council planned to pursue the matter in court.
Council members also reviewed proposed local measures, including a whistleblower incentive program, a right for New Yorkers harmed by AI agents to sue, and requirements for independent third-party checks. These are proposals, not rules that have been adopted. Menin said the hearing was meant to gather testimony and input as the council considers what safeguards the city can enact.
Why it matters
For New Yorkers, the hearing put possible routes to accountability on the table, including a way to sue if an AI agent causes harm. But those measures are still proposals, and the hearing did not establish that the companies can quantify or prevent catastrophic risks.