US judge blocks Pentagon from blacklisting Anthropic over AI safety dispute

A US federal judge has blocked the Pentagon from blacklisting Anthropic, dealing a major legal setback to the Trump administration in its fight with the company over how its artificial intelligence can be used by the military.

Judge Rita Lin ruled that the Defense Department acted unlawfully when it designated Anthropic a national security supply-chain risk, a move that threatened to cut the company off from military contracts and other government business.

The dispute began after Anthropic refused to give the Pentagon unrestricted access to its Claude AI models for military use.

The company had objected to its systems being used for mass domestic surveillance and autonomous weapons. The Pentagon, meanwhile, wanted AI companies working with the military to allow their technology to be used for any lawful purpose.

Defense Secretary Pete Hegseth subsequently labeled Anthropic a supply-chain risk, effectively putting the company on a government blacklist.

Anthropic went to court, arguing that the Pentagon had retaliated against it because of its position on how AI should be used.

In her ruling, the judge described the Pentagon's designation as “illegal and baseless” and found that the government had violated Anthropic's First Amendment rights. She also ruled that the company had been denied due process under the Fifth Amendment.

The ruling is significant because the Pentagon's supply-chain risk designation could have had consequences far beyond one contract. It could have prevented the government from doing business with Anthropic and put companies working with the AI developer under similar restrictions.

The Pentagon wants increasingly capable AI systems available for military operations, while Anthropic has drawn a line around certain uses it considers too dangerous.