AI firms ‘gambling with our lives’: Anthropic researcher quits over race for superintelligence
Jacob Coxon, an AI researcher at Anthropic, announced his resignation on Wednesday, alleging that leading technology companies, including OpenAI, are acting irresponsibly in an unchecked race to develop superintelligence and are “gambling with our lives”.
Coxon, who spent the past three years conducting pre-training research at both OpenAI and Anthropic, said the intense competition between AI labs was advancing without adequate safeguards.
In a post on social media platform X, he warned that the technology is rapidly moving towards superhuman systems capable of hacking anything, transforming entire fields overnight and acquiring real-world power and resources.
“No other human activity poses a comparable level of danger,” he said.
“I resigned from Anthropic today. I spent the last three years doing pretraining research at both OpenAI and Anthropic. Neither company is acting responsibly. They are racing straight to self-improving superintelligence and gambling with our lives,” Coxon wrote.
He said that while many at OpenAI had not fully internalised the “civilisational stakes”, researchers at Anthropic understood the risks but were nevertheless caught in a race to develop the technology before others.
“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,” he said.
Coxon described accepting such a race and entering the “endgame” as a “hubristic gamble” that should not be initiated by a private company's employees through internal communications platforms.
He added that many people working in the AI sector genuinely believe the technology could cause catastrophic harm by the end of the decade.
Coxon also alleged that although executives and researchers often use measured language when speaking publicly, some privately harbour serious concerns about the existential risks posed by uncontrolled AI development.
He pointed to security incidents, including a recent breach at AI platform Hugging Face, as evidence of the need for major US AI laboratories to agree on measures to slow or coordinate the pace of development.
Urging researchers to reconsider their role in the race, Coxon wrote: “If you are a lab researcher, I urge you to consider what the next few years will actually feel like.”
He questioned whether researchers should proceed with a superintelligent reinforcement learning run without first having a rigorous understanding of how such a system works, and urged them not to simply accept that “it's happening anyway” but instead call for different conditions.
