D.C. Circuit upholds Pentagon's Anthropic exclusion

A three-judge panel of the U.S. Court of Appeals for the District of Columbia Circuit ruled 2–1 on September 25, 2026, to uphold the Defense Department's designation of Anthropic PBC and its Claude artificial-intelligence models as a supply-chain risk ABC News.

The Ruling and Its Scope

The appellate court's majority concluded that the Defense Department had "ample support for its conclusion" that continued use of Claude by the department or its contractors presented a national-security risk Politico. The ruling upheld the Pentagon's authority to remove Claude models from Defense Department systems and to prohibit their use in Defense Department contractor-supported work.

The decision means that Anthropic's challenge to the Pentagon's authority under the Federal Acquisition Supply Chain Security Act of 2018 failed on multiple grounds AP News. Anthropic had argued the designation was arbitrary, unauthorized, unconstitutional, and violated due process, but the majority rejected all of those contentions.

Background of the Dispute

The underlying conflict centered on Anthropic's restrictions on certain military uses of Claude and an unresolved contract dispute between the company and the Pentagon. The majority panel treated those circumstances as sufficient support for invoking the Pentagon's supply-chain-risk designation authority.

The designation applies specifically to Defense Department procurement and contractor networks—it does not bar Anthropic or Claude from the entire U.S. federal government CNBC. Other federal agencies retain the ability to use Claude models independently of this ruling.

Industry and Legal Implications

The technology industry expressed concern about the precedent. The Computer and Communications Industry Association noted that the ruling could expand government authority to exclude AI vendors based on perceived national-security risks, even absent criminal conduct or civil penalties CCIA.

The case underscores ongoing tension between AI companies' commercial interests and government security mandates. Anthropic has not yet announced whether it will seek further appeal or challenge, though the company has historically maintained that Claude's safety design and use restrictions align with responsible AI deployment principles MeriTalk.