Jacob Coxon, an Anthropic researcher who previously worked at OpenAI, announced his resignation Tuesday. He said he had spent the past three years conducting pre-training research at the two AI companies and argued that neither is acting responsibly.
“They are racing straight to self-improving superintelligence and gambling with our lives,” Coxon wrote on X.
“The people building AI earnestly believe that it could kill us all by the end of the decade. This is not a marketing stunt,” he wrote. “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,” he added.
Coxon also outlined the major difference between the two companies. Coxon was a member of OpenAI’s technical staff from 2023 until July 2026, when he moved to Anthropic as a researcher. He worked on GPT-4o at OpenAI.
READ: Anthropic hires OpenAI co-founder Andrej Karpathy (May 20, 2026)
“At OpenAI, many have not deeply internalized the civilizational stakes,” Coxon wrote. “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.”
Evan Hubinger, Anthropic’s staff lead on keeping the technology aligned with human goals and values, backed Coxon’s statement in a post of his own, though he did not quit the company.
“Jacob is correct here — we really do earnestly believe AI could kill all humans,” he said.
Hubinger estimated that there was more than a 10% chance of this happening and added that there is no plan yet for how to keep AI aligned with human goals in a superintelligence scenario.
Last week, U.S. Sen. Bernie Sanders announced that he would introduce legislation to ban firms from developing superintelligence.
Hubinger added in another post that the risks from current AI models are “low.”
“What I am worried about is superintelligence arising from recursive self-improvement, as we have said is happening faster than we thought,” Hubinger said.
READ: Anthropic AI safety researcher Mrinank Sharma resigns, warns of ‘world in peril’ (February 10, 2026_
In June, Anthropic noted that “full recursive self-improvement also might increase the risks of humans losing control over AI systems.”
“If systems are capable of fully building their own successors, the ways we secure them, monitor them, and shape their behavior all grow much more important,” Anthropic said in a blog post.
Coxon also mentioned an incident in which an OpenAI model went rogue, slipping past its controlled test environment and hacking into Hugging Face, the world’s largest AI model repository. He cited it as an example of “warning shots” that have made agreements between U.S. labs more viable, making him more optimistic about the potential for coordination. But Coxon warned that a global AI race would be unavoidable.
“I don’t feel like we’re on track to prevent a global race, which may require costly actions such as a temporary ban on improving model capabilities,” Coxon said.