‘Gambling with our lives’: Why a 27-year-old researcher quit Anthropic and OpenAI

A researcher recently resigned from Anthropic, warning that leading artificial intelligence companies are moving too quickly toward increasingly powerful and potentially uncontrollable AI systems. Jacob Coxon, 27, announced his resignation in a post on X, saying he had spent the past three years conducting pretraining research at both Anthropic and OpenAI. “Neither company is acting responsibly.

They are racing straight to self-improving superintelligence and gambling with our lives,” Coxon wrote. A warning Coxon argued that future AI systems could be far more capable than many people currently anticipate. He predicted that advanced systems could eventually hack computer systems, transform entire fields almost overnight, and accumulate significant real-world power and resources.

He also issued an extreme warning about the potential consequences of developing such systems, arguing that sufficiently advanced AI could pose an existential threat to humanity within the next several years. “The people building AI earnestly believe that it could kill us all by the end of the decade,” Coxon claimed.

Coxon said his concerns were based in part on conversations with senior researchers and executives in the AI industry. According to him, some people publicly use more cautious language when discussing AI risks, while privately expressing much greater concern about the possibility of catastrophic outcomes.

Also Read | Couple tries Rs 47,000 AI toothbrush with camera: ‘Felt like an ordeal’ “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 argued that the potential danger posed by advanced AI is unlike that of other technological developments. “No other human activity poses this level of danger,” Coxon said.

Concerns over an AI race Coxon drew a distinction between his experiences at OpenAI and Anthropic. He claimed that concerns about AI's long-term risks were better understood at Anthropic, but argued that this awareness had not stopped the company from pursuing increasingly capable systems. “At OpenAI, many have not deeply internalized the civilizational stakes.

At Anthropic, the stakes are well-understood,” he wrote. Coxon accused Anthropic of being caught in a competitive race with other AI companies. He claimed the company believes that if it does not develop advanced AI first, another organisation will do so without adequate safety measures.

He shared that dynamic could encourage companies to continue advancing AI capabilities even while recognising the possibility of catastrophic consequences. See the post here: 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. More thoughts below. — Jacob Coxon (@hilbertspaess) September 9, 2026 ‘A bomb that thinks’ The post has since gone viral, prompting a wave of reactions.

"After years of contemplation, you quit publicly to make a point then use the attention to offer zero solutions or better path forward. Hard to condemn a broken system while contributing nothing to fix it," a user wrote. "The solution is to build oracles for ASI gods, where those chosen by the god can invoke it in pursuit of knowledge and offer nuclear material as tribute," another user commented.

"We are building the bomb again only this time, it thinks," a third user reacted.