Trump administration illegally retaliated against Anthropic, judge rules

A U.S. district judge ruled that the Trump administration illegally retaliated against Anthropic, the AI startup behind the Claude model, after the company refused to let the Pentagon use its technology for surveillance or autonomous weapons.

The judge, Rita Lin, found that Defense Secretary Pete Hegeth exceeded his authority by designating Anthropic as a “supply‑chain risk” in February—a label normally applied to foreign firms deemed a threat to national security. The designation forced contractors working with the military to boycott Anthropic, prompting the company to sue on grounds that the action was unprecedented, unlawful, and a violation of its First‑Amendment rights. The ruling highlighted that invoking national‑security concerns does not give the government a “blank check” to punish critics, and it underscored the tension between private AI developers and the military’s push to integrate advanced models into defense operations. Anthropic argued its models were not reliable enough for weapons systems and that using them for domestic surveillance would breach civil liberties.

The Pentagon countered that private firms should not be able to limit military use of their technology and expressed concern that Anthropic’s refusal to accept revised contract terms signaled a broader risk to how the technology might be employed. Anthropic welcomed the decision, stating it remains committed to working with the government to harness AI for national security while safeguarding American interests. The company said the government’s actions had harmed its business and stifled public debate about AI’s role in warfare. The judge’s finding of “classic First Amendment retaliation” could set a precedent for how the government engages with tech firms that dissent from military applications, and both the Pentagon and the White House have been contacted for comment on the ruling.

⚡ Effects Interpreter