EXCLUSIVE: Former Anthropic researcher WARNS of ‘AI takeover’.

Warnings from Jacob Coxon, who resigned from Anthropic after working at both Anthropic and OpenAI, claiming companies are racing to self-improving superintelligence while "gambling with our lives."

Coxon stated that AI developers "earnestly believe" the technology could kill us all by the end of the decade, citing superhuman capabilities in hacking, innovation, and resource acquisition as imminent risks.

Anthropic alignment lead Evan Hubinger publicly affirmed the concerns, estimating over 10% chance of AI killing all humans within the next decade due to lack of solved alignment for superintelligence.