A federal appeals court in Washington, D.C., on Wednesday refused to temporarily block the Pentagon’s national security designation of AI company Anthropic, handing the Trump administration a procedural victory as litigation continues. The ruling contrasts with a separate appellate decision that sided with Anthropic in a different challenge.

Anthropic, which develops the widely used Claude AI assistant, contends that Defense Secretary Pete Hegseth exceeded his authority by labeling the firm a national security supply-chain risk after it declined to remove certain safeguards from its products. The designation bars Anthropic from securing Pentagon contracts and could pave the way for a broader government-wide blacklist.

Company leaders have warned the classification could result in billions of dollars in lost revenue and long-term reputational damage.

A panel from the U.S. Court of Appeals for the District of Columbia Circuit rejected Anthropic’s request to pause the designation while the legal dispute proceeds, emphasizing that the order is not a final determination on the merits.

Following the decision, an Anthropic spokeswoman said in a statement that the company remains confident the courts will ultimately find the supply-chain risk label unlawful.

Acting Attorney General Todd Blanche praised the outcome in a social media post Wednesday, calling it a boost for national defense.

"Military authority and operational control belong to the Commander-in-Chief and Department of War, not a tech company," Blanche said, using Trump's new name for the Defense Department.

The case is one of two lawsuits Anthropic has filed challenging Hegseth’s unprecedented action, which followed the company’s refusal to allow its Claude chatbot to be used for U.S. surveillance or autonomous weapons applications, citing safety and ethics concerns.

Hegseth issued the designation under two separate legal authorities, both of which Anthropic is contesting in different courts.

In one of those cases, a federal judge in California on March 26 blocked part of the Pentagon’s action, writing that the government appeared to have unlawfully retaliated against the company over its stance on AI safety.

The designation marked the first time a U.S. firm has been publicly identified as a supply-chain risk under little-known procurement laws designed to protect military infrastructure from foreign interference or sabotage.

Anthropic argues in its lawsuits that the government violated its First Amendment rights by punishing the company for its views on AI safety. It also claims it was denied the opportunity to challenge the designation beforehand, infringing on its Fifth Amendment due process protections.

Court filings from Anthropic assert the government’s actions lack factual support and conflict with prior military praise for the Claude system.

The Justice Department, however, maintains that Anthropic’s refusal to ease its product restrictions could create uncertainty within the Pentagon over how the technology may be deployed and potentially disrupt military operations.

Government attorneys argue the designation was based on the company’s unwillingness to agree to contractual conditions, rather than its public stance on AI safety.

The case in Washington centers on a statute that could expand the blacklist across federal agencies following an interagency review.

Meanwhile, the California litigation addresses a narrower provision that specifically prevents Anthropic from participating in Pentagon contracts tied to military information systems.