US Judge Blocks Pentagon Blacklisting of Anthropic

A U.S. federal judge has blocked the Pentagon from blacklisting artificial intelligence company Anthropic, delivering a major legal victory to the maker of Claude in its escalating dispute with the U.S. military over how AI should be used in warfare and surveillance.

U.S. District Judge Rita Lin issued the ruling on Thursday in a 59-page order, concluding that the Pentagon's decision to designate Anthropic as a national security supply-chain risk was "illegal and baseless."

Lin, who was appointed by former Democratic President Joe Biden, said national security concerns could not be used as justification for retaliating against a company over its policy positions.

"The empty invocation of national security is not a blank check to punish and retaliate against government critics," Lin wrote.

The ruling follows a lawsuit filed by Anthropic in a California federal court challenging Defense Secretary Pete Hegseth's decision to classify the company as a supply-chain risk.

Such a designation can be used by the U.S. government when it believes a company could expose military systems to infiltration, disruption or sabotage by foreign adversaries.

Hegseth's decision came after Anthropic refused to permit the military to use its Claude AI systems for certain forms of surveillance or for autonomous weapons. The Pentagon's move restricted the company's ability to secure some military contracts and threatened significant commercial and reputational damage.

Anthropic executives have warned that the Pentagon's action could cost the company billions of dollars in lost business.

The company welcomed Thursday's ruling, saying it remained committed to cooperating with the U.S. government on the responsible use of artificial intelligence for national security.

"We remain focused on working productively with the government to harness AI for our national security so all Americans benefit from this technology," Anthropic said in a statement.

The Pentagon did not immediately respond to a request for comment.

At the centre of the dispute is a fundamental disagreement over the limits that should be placed on military applications of artificial intelligence.

Anthropic has argued that current AI systems are not sufficiently reliable to be safely deployed in autonomous weapons. The company has also opposed the use of its technology for domestic surveillance, citing concerns over civil liberties and individual rights.

The Pentagon, meanwhile, has argued that private technology companies should not be able to impose restrictions that limit the military's ability to conduct operations.

The Pentagon's designation was particularly significant because it marked the first public designation of a U.S. company as a supply-chain risk under a relatively obscure federal procurement law intended to protect military systems from foreign interference.

In its lawsuit, Anthropic argued that the government had violated its First Amendment rights by retaliating against the company's position on AI safety.

The company also claimed it had been denied due process under the Fifth Amendment because it was not given an adequate opportunity to challenge the supply-chain designation.

Anthropic further argued that the Pentagon's decision was unsupported by evidence and contradicted previous government assessments that had praised Claude's capabilities.

The Justice Department disputed those claims, arguing in court filings that Anthropic's refusal to remove its restrictions could create uncertainty about the Pentagon's ability to use Claude and potentially leave military systems vulnerable to disruption during operations.

Government lawyers maintained that the designation was based on Anthropic's refusal to accept specific contractual conditions rather than on the company's broader views about AI safety.

The legal battle is not yet over. Anthropic is also pursuing a separate case in Washington, D.C., challenging another Pentagon supply-chain risk designation that could potentially prevent the company from obtaining contracts with civilian government agencies.

The latest ruling nevertheless represents a significant setback for the Pentagon as the U.S. government grapples with how rapidly advancing AI technology should be integrated into national defence while balancing operational requirements with safety and civil-liberties concerns.