Former OpenAI and Anthropic Researchers Warn AI Extinction Risk Exceeds 10% This Decade
Original title:Anthropic scientist puts the odds of AI destroying humanity above ten percent this decade
Internal discord over safety and governance is surfacing once again inside top frontier labs. Jacob Coxon, a former pretraining researcher at OpenAI and Anthropic, has resigned while publicly accusing both organizations of consciously pushing forward despite existential risks. In parallel, Anthropic researcher Evan Hubinger placed the odds of a misaligned superintelligence wiping out humanity within the next decade at over ten percent, reflecting a deepening fracture among technical staff.
Why it's worth reading
High-profile resignations and explicit p(doom) assessments from core technical staff highlight escalating friction between commercial scaling imperatives and safety alignment inside leading frontier labs.