Anthropic Researcher Resigns Over AI Superintelligence Risks

Jacob Coxon, a former researcher at both Anthropic and OpenAI, recently announced his resignation on social media. He stated that the leading artificial intelligence companies are rushing toward self improving superintelligence and taking major risks with human safety.

Coxon spent three years working on pre training at the firms. He claimed that upcoming systems will gain real power and resources, and he alleged that industry leaders express severe existential fears privately while softening their tone for the public.

Evan Hubinger, a researcher at Anthropic, responded online to confirm that many builders genuinely believe advanced artificial intelligence poses a severe threat within the decade. Other high profile departures have also hit the sector, including former OpenAI chief futurist Joshua Achiam.

Coxon called for better coordination and suggested that preventing a dangerous race might require temporary bans on capability improvements. Lab staff have previously pushed for government support on pacing agreements to manage automated artificial intelligence development safely.