‘Not a Marketing Stunt’ — Ex-OpenAI and Anthropic Researcher Who Quit Explains Why They Keep Building AI That ‘Could Kill Us All’ Within 10 Years

The people building the world’s most advanced artificial intelligence (AI) systems may be more worried about the technology than they let on publicly. One former researcher says some of the industry’s biggest names privately fear AI could eventually become powerful enough to take out every human on Earth.

Jacob Coxon, a 27-year-old researcher who worked on pretraining at both OpenAI and Anthropic, announced his resignation from Anthropic on Sept. 9 in a thread on X. After three years working at the two AI companies, he said neither is handling the race toward increasingly capable systems responsibly.

“I spent the last three years doing pretraining research at both OpenAI and Anthropic. Neither company is acting responsibly,” Coxon wrote. “They are racing straight to self-improving superintelligence and gambling with our lives.”

Sponsored Content: Own a Piece of the Platform Opening the Private Markets. StartEngine’s latest round is now open to the public.

AI Could ‘Kill Us All’ by 2030

Coxon didn’t frame the risk as a distant science-fiction scenario. He argued that increasingly capable AI systems could arrive much sooner than many people expect.

“The people building AI earnestly believe that it could kill us all by the end of the decade,” he wrote. “This is not a marketing stunt.”

Coxon also claimed that public statements from some executives and researchers don’t fully reflect what they say privately.

“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,” Coxon wrote. “No other human activity poses this level of danger.”

Coxon said the technology’s potential goes well beyond today’s chatbots and productivity tools.

“Do not underestimate the power of this technology,” he wrote. “These will soon be superhuman systems that can hack anything, revolutionize any field overnight, and acquire real power and resources.”

Coxon told The Wall Street Journal that he believes the situation could deteriorate quickly, saying “by the end of next year things could be out of control already.”

Anthropic and OpenAI Face a Race

Coxon also addressed the obvious question behind his decision. If researchers believe advanced AI could pose such a serious threat, why are companies continuing to build it?

His answer differed between the two companies.

“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.”

That race, he argued, creates its own danger.

“Accepting this race and entering the ‘endgame’ is a hubristic gamble that should not be launched from a private company’s Slack,” Coxon wrote. “Attempting to speedrun alignment should require extraordinary confidence that there are no better trajectories available.”

The comments are particularly notable because Anthropic’s own alignment researchers have publicly acknowledged the possibility of catastrophic AI risk.

Evan Hubinger, Anthropic’s Alignment Science Lead, responded to Coxon’s posts on X. Hubinger said he personally believes there is a more than 10% chance AI could kill all humans within the next decade. He also said Anthropic does not yet have a plan to solve alignment for superintelligence.

Pressure Builds Around Frontier AI

Coxon’s resignation comes as concerns about the pace of AI development spread beyond a small group of safety researchers.

More than 1,300 employees from leading technology companies signed a July letter calling on the U.S. government to support pacing frontier AI development.

OpenAI Chief Scientist Jakub Pachocki also warned this month that AI capabilities are advancing faster than researchers’ ability to monitor them.

For Coxon, that gap is the central problem. The technology is advancing rapidly while the systems and safeguards meant to keep increasingly powerful AI under control are still being developed.

His resignation doesn’t mean the AI race is slowing down. It adds another warning from someone who spent years inside two of the companies at its center.

On the date of publication, Caleb Naysmith did not have (either directly or indirectly) positions in any of the securities mentioned in this article. All information and data in this article is solely for informational purposes. For more information please view the Barchart Disclosure Policy here.