Jacob Coxon, a 27-year-old AI researcher, resigned from Anthropic on September 8, 2026, after four months at the company. Previously, he spent three years at OpenAI, where he contributed to the development of GPT-4o.
In his resignation, Coxon warned that the current industry focus on recursive self-improvement could make AI systems uncontrollable by the end of 2027. He argued that unilateral safety restrictions are ineffective and called for global cooperation, specifically involving China, to enforce development limits.
Coxon left the company before his equity fully vested. Anthropic is currently valued in the tens of billions of dollars and is reportedly preparing for a potential IPO, similar to OpenAI.
The researcher's public statement regarding AI safety concerns has gained significant attention, accumulating over 100 million views on X. His departure highlights ongoing internal tensions regarding safety priorities within leading AI labs.
