Anthropic’s Dario Amodei is struggling to swallow a bitter lesson in realpolitik: CEOs moralising about how the military ought to wage wars won’t go down well with the Pentagon.
This follows the Pentagon’s decision to fire Anthropic after it refused to grant the military unrestricted use of its AI technology on ethical grounds. Anthropic’s concern is that frontier AI models aren’t reliable enough to be fully autonomous weapons. To be sure, Pete Hegseth’s subsequent blacklisting of Anthropic from government work over its refusal to relinquish its ethical caveats was heavy-handed.