Anthropic researchers warn out-of-control AI could destroy humans

Several current and former Anthropic researchers are offering a stark warning about AI, suggesting the technology they are developing could destroy humanity in the near future.

Jacob Coxon, a former researcher at both Anthropic and OpenAI, argued Tuesday that neither company is acting responsibly as they race toward superintelligence and accused the AI firms of “gambling with our lives.”

“The people building AI earnestly believe that it could kill us all by the end of the decade,” Coxon wrote in a post on the social platform X, after announcing he had resigned from Anthropic.

“This is not a marketing stunt,” he continued. “If anything, many executives and senior researchers will couch their phrasing in the press to sound sensible – but I hear the same people express fear privately. No other human activity poses this level of danger.”

His comments were quickly echoed by other Anthropic researchers.

Evan Hubinger, who currently heads up alignment science at Anthropic, said “we really do earnestly believe AI could kill all humans ,” adding he thinks there is a greater than 10 percent chance this occurs within the next decade.

He suggested Anthropic is “trying its best” but underscored that researchers don’t have a plan to keep superintelligence aligned with their interests.

Samuel Marks, who leads scalable oversight at the company, similarly said AI developers believe their technology could cause “human extinction” or other bad outcomes in the next few years.

“Why do AI developers continue despite the risk?” he wrote in a post on X . “Due to a mixture of commercial incentives and a belief that they are in a race with other, less responsible AI developers that will abuse the technology or develop it less safely.”

“I work on safety research at Anthropic because I hope my work will reduce the chance of these extinction-level bad outcomes,” Marks added.

It is the latest warning from AI researchers about the risks underlying the increasingly powerful technology they are building.

More than 1,300 employees at frontier AI companies signed on to an open letter in July urging the U.S. government to support an international effort to pace automated AI development.

This effort was ultimately backed by both Anthropic and OpenAI. However, the two companies are also racing one another and fellow AI firms to develop the technology and secure funding, including by going public.

Despite this dynamic, Coxon suggested Tuesday he is “optimistic about the potential for coordination” following “warning shots” from recent incidents in which rogue AI agents gained access to the internet and independently hacked into other companies.