NYC Council questions AI executives on safety as SpaceXAI skips hearing
In short
The New York City Council held a hearing on Monday on AI safety, with policy executives from OpenAI, Meta, Anthropic and Google testifying as the council weighs guardrails proposals. Elon Musk's SpaceXAI was the only company called that did not appear, and the council says it violated a subpoena and will pursue the matter in court. Former Anthropic and OpenAI researcher Jacob Coxon told the council the companies were “extremely reckless” and that AI could end in human extinction.
Read the full story 2 min read
The New York City Council held a public hearing on Monday over artificial intelligence safety, with policy executives from OpenAI, Meta, Anthropic and Google appearing as the council weighs several proposals for guardrails. [ 1 , 2 , 3 ]
Council Speaker Julie Menin asked the companies to quantify the risk their technology poses in a worst-case catastrophic scenario. Morgan Dwyer, OpenAI's head of policy development and operations, who testified virtually along with the other representatives, replied that she did not know and also did not think it mattered. Menin called that “flippant at best” and asked: “If you can't quantify what safety or risk is, how do you know the product is safe?”, comparing it to a pharmaceutical company saying it did not know if a drug would kill people, the New York Post reported. [ 3 ]
The executives also presented the technology's benefits, with Dwyer saying “AI is making it possible to solve economic, scientific and social problems faster than ever,” and pointing to medical research advances and tools for small businesses, according to the New York Post. [ 3 ]
Jacob Coxon, a former researcher at Anthropic and OpenAI, testified that the companies were being “extremely reckless given the stakes.” “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,” he told the council, the New York Post reported. CNBC reported that Coxon quit his job at Anthropic and said on social media that the company and OpenAI were “gambling with our lives.” [ 1 , 3 ]
Elon Musk's SpaceXAI was the only company called to testify that did not appear. Menin said the council had issued a subpoena to Musk the previous week and that the absence was in “direct violation of the subpoena”; she said the council plans to “pursue this matter in court,” CNBC and UPI reported. Menin also said the federal government's proposal that AI will self-regulate “defies all reason,” while council member Oswald Feliz, chair of the Public Safety Committee, said “one of the biggest concerns is that there has been no guardrails created,” UPI reported. [ 1 , 2 ]
CNBC reported that Anthropic, OpenAI and Google agreed to appear only after the council threatened to subpoena them, while Meta confirmed late last month that it would participate; the New York Post also said the companies agreed to testify after being threatened with subpoenas. UPI listed the witnesses as Logan Graham, head of Anthropic's Frontier Red Team; Dwyer; Alice Friend, Google's director of AI and emerging tech policy; and Shane Cahill, Meta's AI policy director for legislation. [ 1 , 2 , 3 ]
CNBC reported that industry researchers warned in September that OpenAI and Anthropic are racing to build models that could potentially cause catastrophic harm, including human extinction, prompting calls for meaningful government oversight. [ 1 ]
Why it matters
The hearing puts the largest AI developers under questioning by a major US city considering its own restrictions, at a moment when researchers and former employees are publicly warning about catastrophic risk. It also sets up a court fight between the council and Elon Musk's AI unit over a subpoena, which could test how far municipal authorities can compel testimony from AI companies.
Key facts
- The New York City Council held a hearing on Monday, Oct. 5, on artificial intelligence safety, with representatives of OpenAI, Meta, Anthropic and Google testifying as the council weighs proposals for guardrails. [ 1 , 2 , 3 ]
- Council Speaker Julie Menin asked the companies to quantify the risk AI poses in a worst-case catastrophic scenario; OpenAI's Morgan Dwyer replied she did not know and did not think it mattered, which Menin called “flippant at best.” [ 3 ]
- Former Anthropic and OpenAI researcher Jacob Coxon told the council the companies were being “extremely reckless given the stakes” and that AI could end in human extinction. [ 1 , 3 ]
- Elon Musk's SpaceXAI was the only company called to testify that did not appear; the council said the absence violated a subpoena and that it plans to pursue the matter in court. [ 1 , 2 ]
- Anthropic, OpenAI and Google agreed to appear only after the council threatened to subpoena them, according to CNBC, and the New York Post also reported the companies agreed to testify after being threatened with subpoenas. [ 1 , 3 ]
- Council member Oswald Feliz, chair of the council's Public Safety Committee, said “one of the biggest concerns is that there has been no guardrails created.” [ 2 ]
- The witnesses included Logan Graham of Anthropic, Morgan Dwyer of OpenAI, Alice Friend of Google and Shane Cahill of Meta, according to UPI. [ 2 ]
Confirmed by several sources
- The New York City Council held a hearing on AI safety on Monday with representatives of OpenAI, Meta, Anthropic and Google appearing to testify. [ 1 , 2 , 3 ]
- The council is weighing proposals for AI guardrails. [ 1 , 2 , 3 ]
- Elon Musk's SpaceXAI did not attend the hearing even though it was called; the council said the absence was in “direct violation of the subpoena” and that it plans to pursue the matter in court. [ 1 , 2 ]
- Jacob Coxon, a former researcher at Anthropic and OpenAI, appeared at the hearing and testified about the risk he says AI poses. [ 1 , 2 , 3 ]
- Julie Menin is the City Council speaker. [ 2 , 3 ]
- Anthropic, OpenAI and Google agreed to testify after the council threatened subpoenas. [ 1 , 3 ]
Still unclear
- Why SpaceXAI did not attend the hearing and how Elon Musk or the company responds to the council's subpoena. The documents report the absence, the subpoena and the council's plan to go to court, but contain no comment from SpaceXAI or Musk.
- What specific AI guardrails proposals the council is weighing and whether any will be passed. The documents say the council is weighing several proposals for guardrails but none describes their contents or the path to a vote.
- The identity and full testimony of a witness named Kokotajlo, who CNBC says called on the AI industry to increase transparency and slow leading model development. Only CNBC mentions him; no other document names him, states his affiliation or describes his testimony.
- Whether the federal government has set up a small committee for AI self-regulation, as council members stated. The claim appears only in UPI's account, attributed to council members, with no federal source cited.
What local media are saying
Timeline, local time
- CNBC reports on the hearing lineup, SpaceXAI's absence and the council's plan to pursue the matter in court. [ 1 ]
- UPI reports that the New York City Council is holding the AI safety hearing on Monday with representatives of OpenAI, Meta, Anthropic and Google. [ 2 ]
- The New York Post reports on the hours-long hearing, including Council Speaker Julie Menin's exchange with OpenAI's Morgan Dwyer and testimony from Jacob Coxon. [ 3 ]