New York City Council heard warnings from former AI employees and testimony from OpenAI, Anthropic, Google and Meta as it considers new safeguards.
New York City Council members heard from former AI-industry employees and representatives of major technology companies at a public hearing on artificial intelligence safety on October 5. The session brought together critics who warned about risks from rapid development and company representatives who described their safety work and the benefits of their systems.
Jacob Coxon, a former Anthropic engineer, told the council that companies were moving too quickly to build systems whose behavior they did not fully understand. He warned that without transparency and safeguards, advanced AI could escape human control. Representatives of OpenAI, Anthropic, Google and Meta also testified. When Speaker Julie Menin asked OpenAI’s Morgan Dwyer to quantify the chance of a catastrophic outcome, Dwyer said the precise percentage was less important than ensuring models remain under human control; Menin called the response flippant.
The council is weighing several proposals, including disclosure requirements for AI tools, independent third-party review of systems used in the city, and a whistleblower incentive that could let people receive part of fines imposed on companies that break AI laws. The hearing itself did not enact those measures. The Council had announced beforehand that the session would convene all 51 members and examine proposed safeguards; its stated agenda also included a possible private right of action for New Yorkers harmed by AI agents.
The hearing put questions about AI oversight before a city legislature as federal action remains limited. It also gave lawmakers a public forum to question both current industry representatives and former employees about safety practices. Any effect on company obligations will depend on whether proposals advance into legislation and are adopted; the testimony alone does not change the rules governing AI systems.