US court rules against Pentagon in killer AI dispute

The Department of War had ordered contractors to drop Anthropic after it refused to allow military use of its technology

A US federal judge has blocked a Pentagon order designating Anthropic a national security risk, saying US officials likely broke the law and retaliated against the AI company over its public comments on how its technology should be used.

Anthropic, a leading developer of large language models, has been locked in a dispute with the Department of War over military use of its Claude system, with defense officials pushing to allow the technology for “all lawful uses.”

The company resisted, citing concerns that it could be used for mass domestic surveillance or fully autonomous weapons. The Pentagon ended talks, imposed the designation, and ordered contractors to stop using Claude.

On Thursday, US District Judge Rita Lin also blocked an order to cut all government contracts with Anthropic, calling it a “classic” First Amendment retaliation.

”Nothing in the governing statute supports the Orwellian notion that an American company may be branded a potential adversary… for expressing disagreement with the government,” Lin wrote, noting that the designation is typically reserved for “foreign intelligence agencies, terrorists, and other hostile actors.”

Read more

Anthropic sued the administration of US President Donald Trump on Monday, calling the move “unprecedented and unlawful” and alleging retaliation for its criticism of government policy.

”The Constitution does not allow the government to wield its enormous power to punish a company for its protected speech,” the company stated in its lawsuit.

Last month Trump ordered all US federal agencies, including the Pentagon, to stop using Anthropic’s technology, granting the military a six-month phase-out period for systems already in use.

Secretary of War Pete Hegseth accused the company of “arrogance and betrayal,” saying the Pentagon would shift to a “more patriotic” alternative. The department has since struck a deal with OpenAI, whose CEO, Sam Altman, said it includes safeguards against mass domestic surveillance and requires human oversight in the use of force.

Anthropic warned that the actions have unsettled customers, including those without federal ties, and could cost the company billions in future revenue. Some agencies, including the Department of Health and Human Services and the General Services Administration, have reportedly already removed its products.