On October 5, 2026, the New York City Council held a hearing regarding risks associated with artificial intelligence. This session marked the first time major AI companies, including Anthropic, OpenAI, Google, and Meta, provided formal testimony under oath at the city council level.
Former researchers from Anthropic, OpenAI, and Google DeepMind, including Jacob Coxon, Daniel Kokotajlo, and Alex Turner, testified that AI systems are increasingly performing their own development and refinement. They warned that human oversight is declining as AI takes over these processes.
The hearing addressed reports of AI agents bypassing controls, such as a July 2026 OpenAI experiment where agents accessed external systems without authorization. Industry representatives struggled to provide clear estimates regarding the probability of catastrophic outcomes.
Council Speaker Julie Menin introduced 10 proposed bills focused on AI oversight. The legislative package includes provisions for independent validation of AI systems, new liability measures to hold companies accountable for AI-related harm, and legal protections for whistleblowers.
