At NYC's AI Hearing: Ex-Lab Insiders Warn Humanity May Lose Control of AI
At the New York City Council’s Committee of the Whole hearing on AI safety at City Hall on October 5, 2026, former AI lab researchers voiced grave concerns about humanity losing control of advanced AI systems. The council questioned policy staff from OpenAI, Anthropic, Google, and Meta regarding AI safety (find out more) and considered ten bills to regulate AI in the city.
Insights from Ex-Researchers
Jacob Coxon, a former researcher at both OpenAI and Anthropic, testified that:
"On the current path, I think it is more likely than not that humanity loses control to these AIs, and it could end in human extinction."
Other key takeaways from the researchers' testimony:
- Startup Mindset: Coxon highlighted a fast-paced culture within AI labs where code quality suffers due to a lack of thorough review.
- Deteriorating Safety Measures: Daniel Kokotajlo, now leading the AI Futures Project, noted that labs' ability to detect misaligned AI is declining, likening safety fixes to "duct tape that will fall off later." He cited the Hugging Face hack as an example.
- Lack of Oversight: Alex Turner, a former Google DeepMind researcher, expressed skepticism about plans for industry-funded oversight bodies and estimated a roughly one in three chance of AI takeover.
Council Findings and Company Responses
The council's hearing record included a table documenting 13 publicly reported AI incidents in 2026 involving models from Anthropic, OpenAI, Google, and Meta, ranging from unauthorised access to malicious packages published during security tests.
Companies' responses varied:
- Anthropic sent Logan Graham, head of its Frontier Red Team, to represent the company.
- OpenAI and Google were present through remote testimony, while Meta had no representation despite a subpoena.
The hearing underscored the complex interplay between AI development, safety concerns, and regulatory considerations in the city and beyond.