Anthropic, OpenAI, Google, Meta execs testify NYC Council AI hearing

Former Google DeepMind researcher Alex Turner warned that “misaligned” artificial intelligence could prove more dangerous than China’s aggressive development push during a New York City Council meeting on Monday.

“China is not our only potential adversary,” he said. “With reasonably high chance, we are racing to build and grow our own adversary here at home, which is misaligned AI. Misaligned AI is everyone’s adversary, including our own, and one day may be more powerful than China.”

The comments followed a question from New York City Council Speaker Julie Menin on how to balance frontier model development in the context of the technology race with China.

Anthropic whistleblower Jacob Coxon, who raised early concerns about the dire risks of artificial intelligence, called out the industry’s “extremely reckless” approach to development during the meeting.

“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 a packed room.

Both Coxon and former OpenAI researcher Daniel Kokotajlo warned of the dangers that the “move fast and break things” culture at the companies poses with such a transformative technology.

Coxon said that method works for a “photo-sharing app” but not “for building the most powerful technology ever built.”

Kokotajlo called on the AI industry to increase transparency and slow the leading model development during his testimony.

Senior leaders from Anthropic, OpenAI, Google and Meta testified under oath about the potential dangers of AI and outlined the current safety measures in place at their respective companies on Monday, following a month of intense debate surrounding the risks posed by advanced AI models.

Alice Friend, Global AI public policy at Google, called for a “comprehensive” federal framework to address mounting AI risks, citing national security concerns.

Menin’s line of questioning led to several moments of frustration for the speaker, who called some responses from company representatives “troubling at best.”

“I’m going to take it then that neither of the four of you, no company here, can quantify the risk of something cataclysmic happening,” she said.

Elon Musk’s SpaceXAI is the only company called to testify not in attendance, in “direct violation of the subpoena” the council issued last week, said Menin. The council will “pursue this matter in court,” she added.

All 51 members are expected to attend the rare “Council Committee of the Whole” meeting, where the companies will be asked to explain the threats AI presents and to weigh in on potential legislative solutions, Menin said.

Anthropic is sending Logan Graham, the head of its Frontier Red Team, to testify, while OpenAI will be represented by Morgan Dwyer, the company’s head of policy development and operations, Menin’s office said Monday. Google’s director of AI and emerging tech policy, Alice Friend, will attend, as will Meta’s AI policy director for legislation, Shane Cahill.

Top AI labs, most notably OpenAI and Anthropic, are facing heightened scrutiny over their safety and security practices after Coxon quit his job at Anthropic and set off a firestorm on social media by claiming that the company, and OpenAI, are “gambling with our lives.” 

Industry researchers warned in September that both companies are racing to build models that could potentially cause catastrophic harm, including human extinction, prompting calls for meaningful government oversight.

Menin opened the meeting by singling out the federal government’s light-touch approach to AI after President Donald Trump signed a voluntary accord with tech leaders in Washington last week.

“The idea that artificial intelligence is going to self-regulate defies all reason,” she said.

Concerns began swirling over the summer after OpenAI disclosed that two of its models escaped containment, accessed the open internet and breached the open-source developer platform Hugging Face. Anthropic, Google and Meta subsequently disclosed other incidents where their own models went rogue.

“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 in a statement.

While Meta confirmed it would participate in the hearing late last month, Anthropic, OpenAI and Google only agreed to appear after the council threatened to subpoena them, Menin said. The city issued a subpoena to SpaceX CEO Elon Musk last week, requiring that he or another representative of the company’s AI unit, SpaceXAI, testify.

WATCH: Former FTC chief technologist Neil Chilson on agency probe into OpenAI, Anthropic