The White House tried to slap Anthropic with a national security stink bomb, and a federal judge has told them to sling their hook.
A US federal judge has blocked the Trump administration from punishing Anthropic after the AI start-up refused to allow its technology to be used without limits in warfare.
Northern District of California judge Rita Lin said the “financial and reputational harm” from the government branding Anthropic a supply chain risk could “cripple” the company. She temporarily barred US agencies from enforcing the designation.
That supply chain risk label is normally aimed at outfits tied to foreign adversaries, not a homegrown firm arguing about what its models should be allowed to do.
The decision is expected to be appealed, landing weeks after Pentagon talks with Anthropic collapsed when it would not support lethal autonomous weapons and mass surveillance.
Anthropic’s Claude model had been used by the military in classified work, including in the Iran war and the capture of Venezuelan leader Nicolás Maduro.
After the public scrap, Donald Trump labelled Anthropic a supply chain risk, a move the company says has never been used on an American business.
The administration barred civilian agencies from doing business with Anthropic, while US defence secretary Pete Hegseth suggested vendors wanting US military contracts must end commercial relationships with the $380bn start-up.
Anthropic sued, saying the designation would cause enormous damage and that it was being unlawfully targeted for criticising the administration.
Several large tech companies filed briefs backing Anthropic, arguing the designation was “causing immediate and substantial harm to the technology industry”.
Lin said it was the Pentagon’s “prerogative to decide what AI product it uses” but that the attempt to brand Anthropic a supply chain risk was “likely contrary to law, arbitrary and capricious”.
“Nothing in the governing statute supports the Orwellian notion that an American company may be branded a potential adversary and saboteur of the US for expressing disagreement with the government,” Lin said.
The administration has seven days to appeal the injunction, and it will not take effect until that window closes.
Anthropic said it was “grateful to the court for moving swiftly”.
“While this case was necessary to protect Anthropic, our customers and our partners, our focus remains on working productively with the government to ensure all Americans benefit from safe, reliable AI,” it added.
The injunction means the government does not need to immediately end its commercial partnerships with Anthropic and, at a hearing on Tuesday, Pentagon lawyers said contractors could still use it outside defence work.
