Former Anthropic developer warns AI firms are risking human extinction

Jacob Coxon, a former researcher at OpenAI and Anthropic (Anthropic), has resigned, accusing both companies of irresponsible practices in the race to develop self-improving superintelligence. Coxon warned that these systems could soon possess the capability to breach networks and accumulate resources, noting that internal discussions often acknowledge the risk of human extinction by the end of the decade.