An Anthropic researcher's resignation letter has done what corporate PR rarely can: gotten AI's biggest names and the Pentagon to publicly disagree with each other.
Jacob Coxon left Anthropic this week and posted his reasoning on X, warning that AI companies are racing ahead without enough caution. The post did not stay contained to one platform for long. Senior scientists at rival labs backed his central claim, that advanced AI could pose an existential threat to humanity. A senior Pentagon official pushed back just as publicly, rejecting the extinction framing outright.
That split matters more than the resignation itself. When the people building frontier models agree the risk is real, but the institution deciding how to deploy that technology in defense systems does not, the disagreement stops being academic and starts shaping procurement and policy. Right now, those two camps aren't so much debating as talking past each other.
Extinction warnings from AI insiders have become a genre of their own. What's new here is who's shouting past whom.