A federal judge on Thursday issued a temporary halt to the Pentagon’s ban on AI company Anthropic, marking a significant development in the company’s legal battle with the military over artificial intelligence safety concerns.
The artificial intelligence firm behind the Claude chatbot filed a lawsuit in California federal court, claiming Defense Secretary Pete Hegseth exceeded his authority by labeling Anthropic as a national security supply-chain threat. This designation allows the government to identify companies that could potentially expose military operations to enemy interference or sabotage.
The company argued that federal officials violated its First Amendment free speech protections by taking retaliatory action against its stance on AI safety protocols. Anthropic also claimed the government denied its Fifth Amendment due process rights by failing to provide an opportunity to challenge the designation.
U.S. District Judge Rita Lin, who was appointed by former Democratic President Joe Biden, sided with the company in her 43-page decision, though she delayed implementation for seven days to allow the administration time to file an appeal.
The Defense Secretary’s unusual action came after Anthropic declined to permit military officials to utilize its Claude AI system for domestic surveillance operations or autonomous weapon systems. The blacklisting prevented Anthropic from pursuing certain defense contracts, with company leadership estimating potential losses in the billions of dollars plus significant damage to their reputation.
The AI company maintains that current artificial intelligence technology lacks the reliability necessary for safe deployment in autonomous weapons systems and argues that domestic surveillance programs violate civil liberties. Pentagon officials counter that private corporations should not have the authority to limit military operations.
In her Thursday decision, Judge Lin determined that the administration’s response appeared motivated by a desire to penalize Anthropic rather than address legitimate national security concerns.
“The record supports an inference that Anthropic is being punished for criticizing the government’s contracting position in the press,” Lin wrote.
“Punishing Anthropic for bringing public scrutiny to the government’s contracting position is classic illegal First Amendment retaliation,” the judge added.
Anthropic spokesperson Danielle Cohen expressed satisfaction with the court’s ruling.
“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,” Cohen said in a statement.
This marks the first instance of an American company being publicly classified as a supply-chain risk under an obscure procurement law designed to shield military systems from foreign interference.
Anthropic’s March 9 legal filing characterized the decision as unlawful, lacking factual support, and contradicting the military’s previous positive assessments of Claude.
The Justice Department responded that Anthropic’s unwillingness to remove usage restrictions could create operational uncertainty for the Pentagon regarding Claude’s deployment and potentially compromise military systems during active operations, according to court documents.
Government attorneys maintained that the designation resulted from Anthropic’s rejection of contractual terms rather than its position on AI safety issues.
The company faces a separate legal proceeding in Washington, D.C., challenging another Pentagon supply-chain risk designation that could result in exclusion from civilian government contracts.
