Anthropic faces a high-stakes showdown with the Pentagon over whether the Defense Department can use its AI systems for any lawful purpose, as a deadline looms to revise the company’s usage rules. Defense Secretary Pete Hegseth has threatened to designate Anthropic a “supply chain risk” or invoke the Defense Production Act if it refuses, a move that could freeze the startup out of federal work and ripple across contractors. CEO Dario Amodei says the company won’t permit uses tied to fully autonomous weapons or domestic mass surveillance, framing the dispute as a test of its safety commitments even as it juggles investor expectations around a towering $380 billion valuation. The standoff highlights mounting tension between military imperatives and Silicon Valley’s risk guardrails, drawing rare public support for Anthropic from OpenAI’s Sam Altman and sparking solidarity letters across the industry. A punitive outcome could chill private-sector participation in defense AI just as the Pentagon accelerates efforts to reduce policy “constraints” on model use.
Related articles:
Pentagon cancels $10 billion JEDI cloud contract
NIST AI Risk Management Framework
