Court blocks Pentagon blacklist of Anthropic

A US federal judge ruled that the Pentagon unlawfully blacklisted Anthropic as a supply-chain risk and ordered the government to unwind restrictions tied to that designation. The ruling says the government retaliated after Anthropic refused to let Claude be used for lethal autonomous warfare and mass surveillance.

The immediate effect is practical: federal agencies and contractors cannot treat Anthropic as a banned supplier on that basis. The judge said the government had not shown that Claude posed a distinct technical risk, and that a public policy position about contract terms is not the same as sabotage or hidden access to a system.

The case matters beyond one vendor. Frontier AI companies increasingly sell into defense and government customers while keeping safety rules around weapons, surveillance and autonomy. The decision narrows how far the government can go when a supplier’s AI-use policies clash with procurement priorities.

The administration can still appeal, and the ruling does not require any agency to buy Anthropic products. It does, however, remove a broad blacklist that could have shut Claude out of government and contractor work.