Anthropic Researcher Resigns, Warns Labs Are Gambling with Lives

A senior pretraining researcher at artificial intelligence firm Anthropic, Jacob Coxon, has abruptly resigned, issuing a public warning that leading tech laboratories are gambling with human lives in their race toward self-improving superintelligence. The resignation follows intense pressure within the artificial intelligence sector surrounding commercial deployment speeds and safety testing protocols. Coxon stated that internal safety guardrails are increasingly sidelined to maintain a competitive advantage against rival firms, noting that builders privately acknowledge existential threats. The departure adds to growing friction inside major American AI companies regarding the oversight of frontier models. Coxon advocates for extreme measures, including a potential temporary ban on improving model capabilities to prevent a global race. Executives at Anthropic and other leading facilities face mounting scrutiny from policy groups over the pacing of public releases. The resignation letter emphasizes an urgent need for independent regulatory oversight before next-generation models scale further.