Former AI researcher: ‘Not enough time’ for employees to consider safety of models

Days after he resigned from Anthropic, former AI researcher Jacob Coxon said Sunday there is “not enough time” for employees to consider the risks of the technology they are working on.

During an interview on ABC’s “ This Week ,” Coxon said safety risks are “definitely not top of mind” for employees of AI firms.

“If you’re working on the tech, then you’re sort of head down trying to do coding, programming. There’s not enough time to be constantly thinking about the risks,” Coxon told host George Stephanopoulos.

Coxon, who worked at Anthropic and OpenAI, left the former company earlier this week and said both companies were “gambling with our lives.”

Coxon wrote Tuesday on social platform X , “The people building AI earnestly believe that it could kill us all by the end of the decade.”

The former AI researcher added that at OpenAI, many employees “have not deeply internalized the civilizational stakes” of the rapid development of AI models.

“At Anthropic, the stakes are well-understood, but they are locked in a race to get there first — they believe no one else will act responsibly, so they must do it themselves, despite the risk,” Coxon argued.

Multiple lawmakers expressed alarm at Coxon’s warning, urging their colleagues to pass safeguards on the technology.

House Minority Leader Hakeem Jeffries (D-N.Y.) said Sunday he supports addressing risks posed by AI. But the New York Democrat said his GOP counterparts have “abandoned their responsibility” on the issue by canceling the final two weeks of the lower chamber’s September session.

“We should take decisive action now,” Jeffries said on “This Week.”

Anthropic CEO Dario Amodei said Saturday he supports “pacing the rate of capabilities advancement” for AI models to allow risk prevention measures to catch up.

“We must slow the pace at which we improve the capabilities of AI models. Progress will still seem fast, and we must make wise use of the time we gain,” Amodei wrote in an essay on his website.

Coxon agreed with his former boss, saying Amodei calling for a slowdown in AI development is a “substantial step forward.”

He told Stephanopoulos, “I think it’s great. It’s very exciting,” adding American companies should also work with their Chinese counterparts on risk prevention.