SAN FRANCISCO — US District Judge Rita Lin granted Anthropic a temporary injunction on Tuesday to pause the Department of Defense's supply-chain-risk designation against the artificial intelligence company. Lin stayed the Pentagon's punitive measures for one week.

Lawyers for Anthropic and the Department of Justice appeared before Lin in San Francisco to argue Anthropic's request for an injunction. "The US Department of Defense appears to be illegally punishing Anthropic for trying to restrict the use of its AI tools by the military," Lin said during the hearing.

The dispute arose after the government applied a supply-chain-risk designation to Anthropic following the company's push for limitations on how its AI could be used by the military. Anthropic has filed two federal lawsuits alleging that the Trump administration's decision to designate the company a security risk amounted to illegal retaliation.

"It looks like [the department] is punishing Anthropic for trying to bring public scrutiny to this contract dispute, which of course would be a violation of the First Amendment," Lin said. She described the supply-chain-risk designation as a powerful authority typically reserved for foreign adversaries, terrorists, and other hostile actors.

Defense Secretary Pete Hegseth posted on X: "Effective immediately, no contractor, supplier, or partner that does business with the United States military may conduct any commercial activity with Anthropic." Lin noted that determining whether Anthropic is an appropriate vendor for the Department of Defense is the role of Hegseth, but said it is up to her to determine whether he violated the law by taking steps beyond canceling Anthropic's government contracts.

The Department of Defense argued that it followed procedures and determined Anthropic's AI tools could no longer be relied upon to operate as expected during critical moments. The department asked Lin not to second-guess its national security assessment and said it is working to replace Anthropic technologies over the coming months with alternatives from Google, OpenAI, and xAI.

"It is troubling that the security designation and directives limiting use of Anthropic's AI tool Claude by government contractors don't seem to be tailored to stated national security concerns," Lin said. She wrote that the government's designation of Anthropic as a supply-chain risk is likely unlawful and arbitrary and capricious. Michael Mongan, a WilmerHale attorney representing Anthropic, said it was "extraordinary for the government to go after a 'stubborn' negotiating partner with the designation."

Lin noted that President Donald Trump and Secretary Hegseth described Anthropic as 'woke' and referred to the company as composed of 'left-wing nut jobs' rather than citing security concerns. Anthropic sued the Department of Defense and several other agencies earlier this month after Trump publicly criticized the company and Hegseth designated it a supply-chain risk.

The temporary injunction means that Anthropic's AI tools, including Claude, will continue to be used by government agencies and by outside companies working with the military until the lawsuit is resolved.