Anthropic Researcher Jacob Coxon Resigns Over AI Safety Fears

In a post on X, researcher Jacob Coxon said he resigned from Anthropic on September 9 over concerns about advanced AI systems. He accused Anthropic and OpenAI of racing toward self improving superintelligence without adequate safeguards.

Coxon spent three years conducting pretraining research at OpenAI and Anthropic. He warned that future systems could gain strong cyber capabilities, accelerate research and acquire resources, while developers may lack a rigorous understanding of how the systems operate.

Coxon called for researchers to reconsider work on superintelligent reinforcement learning systems. He also supported coordination and pacing agreements among United States AI labs.