Anthropic Researcher Resigns Over Fears of AI Risks

Anthropic researcher Jacob Coxon has resigned from the company, citing fears that AI labs are developing systems that could pose an existential threat to humanity within the next few years. Coxon, who previously worked at OpenAI, accused leading AI companies of “gambling with our lives” in the race to develop superintelligence.

In a post on X, Coxon said he had spent the past three years researching pretraining at both OpenAI and Anthropic. He argued that neither company was acting responsibly and that both were moving rapidly toward self-improving superintelligence.

Coxon warned that increasingly powerful AI systems could soon be capable of hacking almost anything, rapidly transforming entire fields and acquiring real-world power and resources.

The 27-year-old former Anthropic researcher said some people developing AI genuinely believe it could kill everyone before the end of this decade. He also claimed that he had personally heard senior AI researchers and executives express similar fears in private conversations.

According to Coxon, these concerns are not simply a marketing strategy. He said many executives and senior researchers publicly frame their views in more measured language, while privately acknowledging the potential dangers of advanced AI.

Coxon also criticized Anthropic for continuing its pursuit of advanced AI despite understanding the potential risks. He said the company believes it must win the race because other companies may not act responsibly.

He argued that the scale of the potential danger makes AI development unlike any other human activity in terms of risk.