Appeals court upholds Pentagon’s Anthropic supply-chain risk designation

A divided federal appeals court in Washington upheld the Pentagon’s designation of artificial intelligence company Anthropic as a supply-chain risk, allowing restrictions on its Claude models to remain in place.

In a 2-1 ruling Friday, the U.S. Court of Appeals for the District of Columbia Circuit rejected Anthropic’s arguments that the Defense Department’s action was arbitrary, unauthorized and unconstitutional. Judge Gregory Katsas wrote that the department had “ample support” for concluding that Claude’s continued integration into Pentagon systems, including through contractors, posed a national-security risk covered by federal law. Judge Neomi Rao joined the majority, while Judge Karen LeCraft Henderson dissented.

The dispute arose after Anthropic declined to remove safeguards restricting Claude’s use for fully autonomous weapons and domestic surveillance. The designation prevents the military from using Anthropic’s models and bars defense contractors from using them in work for the department.

The Pentagon relied on two separate legal authorities for its actions against Anthropic. A federal judge in San Francisco previously ruled that a parallel designation was unlawful, but Friday’s decision upheld the other designation.

Anthropic said it respectfully disagreed with the ruling and was considering further review. The appeals court delayed the decision from taking immediate effect, giving the company time to seek rehearing by the panel or the full D.C. Circuit. Anthropic could also petition the U.S. Supreme Court.