AI researcher Jacob Coxon resigned from US-based AI safety company Anthropic in early 2025 after just four months, citing concerns that the people building artificial intelligence "earnestly believe that it could kill us all by the end of the decade." A senior Anthropic researcher, Evan Hubinger, publicly agreed, estimating a greater than 10% chance of an AI-caused human extinction within ten years. The resignations have intensified debate among scientists and policymakers about slowing AI development and strengthening human oversight, with proposed catastrophic scenarios ranging from AI-engineered bioweapons to social collapse driven by mass unemployment and misinformation.