After a decade of failed legislation and a string of departures, Washington is finally starting to care about AI safety

Jacob Coxon, who has worked at OpenAI and Anthropic, recently left Anthropic, warning that the two companies are “gambling with our lives” in the race to superintelligence. Anthropic alignment science team lead Evan Hubinger publicly backed the claim within hours, estimating the probability of human extinction within ten years at over 10%. Similar warnings from Hinton, Leike, Sutskever, Sharma, and others had not led to legislation.