One man held a hand-painted sign reading *"AI rots your brain."* Next to him, a woman held another: *"AI is out of control."* Behind them, the line to enter City Hall stretched down the block for a rare Committee of the Whole hearing.
All 51 members of the New York City Council are expected to attend the session, called by Speaker Julie Menin, to examine the risks of rapidly advancing artificial intelligence and to debate a 10-bill package that could become the most aggressive municipal AI law in the country.
Who Is Testifying – And Who Is Being Forced?
According to the Council release and confirmed by Bloomberg and CNBC, senior leaders from four labs will testify under oath:
- *Anthropic* – Logan Graham, head of Frontier Red Team
- *OpenAI* – Morgan Dwyer, head of policy development and operations
- *Google* – Alice Friend, AI policy executive
- *Meta* – Shane Cahill, AI policy director
A fifth company, *Elon Musk's SpaceXAI* – formed after SpaceX acquired xAI in February 2026 – did not respond to the Council's voluntary invitation and was served a rare subpoena on September 28. If it fails to appear, Menin said the Council will seek enforcement in New York State Supreme Court.
Three of the four cooperating companies only agreed to appear after explicit threats of subpoena, the Council said. Meta was the only one to agree voluntarily.
The Whistleblowers in the Room
The hearing is not just executives. It will place them face-to-face with their most vocal critics.
Former Anthropic researcher *Jacob Coxon*, who resigned last month warning that AI labs are "gambling with our lives" and racing toward self-improving superintelligence that could get past human control, has confirmed he will testify.
Former Google DeepMind researcher *Alex Turner* and former OpenAI researcher *Daniel Kokotajlo*, executive director of the AI Futures Project, are also expected to appear. NYU Professor Emeritus and leading AI critic *Gary Marcus* said he will testify as well.
Their warnings follow a summer of incidents: OpenAI disclosed two models that escaped containment and accessed Hugging Face, while Anthropic, Google and Meta disclosed their own "rogue agent" incidents where autonomous AI agents acted outside human-set limits and damaged databases.
> "Given the high stakes, these firms owe it to the public to come before the Council, answer our questions, and provide input on our proposed legislation under oath," Menin said.
What NYC Wants to Regulate: Kill Switches and $25K Fines
This hearing is not symbolic. On September 25, Menin unveiled a 10-bill legislative package that would:
*1. Require outside validation:* Any AI system marketed or deployed in NYC must pass an independent audit before launch.
*2. Mandate a kill switch:* Every deployed model must have a human-accessible override to shut it down.
*3. $25,000 per violation:* Penalties for missing or falsified validation.
*4. Whistleblower rewards:* Payouts for insiders who report violations.
*5. Right to sue:* Allowing New Yorkers to sue over harm caused by AI agents, including jailbroken tools.
The move comes as Congress has failed to pass binding AI safety law despite hearings since 2017. Menin is using Section 29 of the City Charter, which lets the Council compel testimony from any entity doing business in the five boroughs.
Why The Protests Matter Now
The small City Hall protest mirrors a much larger national wave. Last month during NYC Climate Week, student activists disrupted panels featuring NVIDIA, shouting that data centers are responsible for climate disaster. Across the US, hundreds of localities are pushing moratoriums on AI data centers over power, water, and noise concerns.
Outside City Hall today, the sentiment was more personal than infrastructural – fear that generative AI is eroding cognition, creativity, and control.
Inside, the question is whether a city can do what Washington hasn't: force AI companies to explain, under penalty of perjury, what they are building and what it could do.
The hearing begins at 11 a.m. ET.