Former Anthropic researcher Jacob Coxon and representatives from Anthropic, Google, OpenAI, and Meta testified at a New York City Council hearing on AI safety
Former Anthropic researcher Jacob Coxon warned that the race to develop artificial intelligence poses risks that the industry itself …
Bloomberg
Context & Ripple Effects
Jacob Coxon’s testimony follows his September departure from Anthropic over fears that labs were building systems they could not control, and his call for industry-wide international coordination on advanced AI risks. His appearance moves that critique from research-lab debate into a municipal oversight forum shared with Anthropic, Google, OpenAI and Meta.
The hearing also extends a policy conversation that reached Congress in 2023, when leading AI researchers and executives warned lawmakers about serious potential harms. It gives New York City Council members a public record of how major developers frame safety, oversight and responsibility.
First-order effects
Anthropic, Google, OpenAI and Meta must place their AI-safety positions before New York City Council members under the same scrutiny as Coxon’s account of risks inside the development race.
Coxon gains an institutional venue for the concerns that prompted his exit from Anthropic, turning an individual researcher’s warning into formal legislative testimony.
Second-order effects
New York City lawmakers gain competing evidence from a former lab researcher and the companies he criticized, sharpening the basis for city-level safeguards and accountability proposals.
Divergent approaches to regulation become more visible: Sam Altman has said OpenAI and Anthropic have fundamentally different views on AI regulation and risk tolerance, making a unified industry response less likely.
Third-order effects
If local hearings become a recurring oversight mechanism, frontier AI firms will face a more fragmented governance environment in which public testimony and operational assurances matter alongside voluntary safety commitments.
The episode points toward a state-compatible AI-lab model, with elected bodies seeking direct leverage over developers as concerns move from expert warnings to concrete oversight demands.
The trend:AI safety governance is shifting from broad warnings by researchers toward direct, multi-company scrutiny by local lawmakers seeking enforceable accountability.
At today's NYC Council hearing, @OpenAI just claimed they supported my RAISE Act to regulate dangerous AI models. I was there—they actively pushed to water down the bill's safety requirements. It's another example of why we can't trust these companies to regulate themselves.
Today's @NYCCouncil Committee of the Whole brings all 51 Councilmembers together to question @AnthropicAI, @OpenAI, @Google and @Meta under oath. My bill would give New Yorkers the right to take legal action against AI companies when foreseeable misuse of their technology causes …
Without whistleblowers, the world would not have learned the full extent of the potential dangers of AI. We thank whistleblowers, including Jacob Coxon who was a researcher at both Anthropic and OpenAI, for bravely testifying today at City Hall. Photo credit: William Alatriste/ @…
NYC city council holding a big hearing today on AI safety with reps from the big labs, except SpaceXAI, which was subpoenaed. City council chair says they “are pursuing that subpoena in court”
SpaceX / Musk received a subpoena to testify at this hearing... www.cnbc.com/2026/10/05/a... “the Council may seek judicial enforcement in New York State Supreme Court if SpaceXAI doesn't comply with the subpoena.”
The @NYCCouncil held a first-in-the-nation hearing to question @claudeai, @openai, @google, and @meta about the societal risks posed by AI. My bill, Int. T2600-2026, would create a private right of action, allowing New Yorkers to sue AI companies when foreseeable misuse of their …
@NYCCouncil is hosting a hearing and introducing a package of legislation aimed at regulating AI as it progresses in our city. At a time when the federal government is lacking the oversight and seriousness needed to deal with this, New York City is stepping up to lead. [video]
Asterisk Magazine, which NY Mag partnered with to publish this cover story, is an Effective Altruism org and the writers on this piece have deep ties to EA.
There are no better chroniclers of the culture of the AGI era than @clarabcollier and @chengela, and this piece is a characteristically excellent illustration of that https://nymag.com/...
Imagine if Leading the Future sponsored a New York Magazine cover story and the (rightful!) outrage that would ensue, but these EA orgs and EA money is flooding the legacy journalism industry and there's effectively zero meaningful backlash or media criticism. It's appalling
Not only that, but Asterisk is funded by Coefficient Giving (formerly Open Philanthropy). It really is remarkable that @NYMag didn't disclose the ties that one of its co-authors on the piece has https://x.com/...
Jacob Coxon quit Anthropic after he came to believe that AI could pose a risk to humanity. Shortly, he will testify at a City Council hearing, along with fellow industry whistleblowers Daniel Kokotajlo and Alex Turner, on the risks of accelerating AI progress. https://nymag.com/.…
THIS. IS. Unbelievable. I am absolutely speechless on this one. There is NO disclosure in the @NYMag piece other than it saying it is a collaboration with Asterisk, even though if you click through to Asterisk it takes some doing to figure out their funding and ties to EA. And wh…