Anthropic Blacklisting Upheld by Appeals Court

Anthropic’s blacklisting remained in place on April 8, 2026, when a three-judge panel of the U.S. Court of Appeals for the District of Columbia Circuit denied the company’s emergency motion to pause the Pentagon’s supply chain risk designation. The court ruled that national-security concerns outweighed the financial harm to Anthropic and accelerated the timetable for resolving the case.

Appeals Court Denies Stay, Advances Merits Hearing

The D.C. Circuit’s denial was interlocutory and did not address the merits of Anthropic’s challenge. Instead, the court prioritized national-security interests over the company’s financial losses, describing the harm as primarily financial. The panel ordered an expedited merits proceeding to resolve the underlying dispute.

The court framed the issue as weighing a contained financial injury to a private company against the government’s need to control access to sensitive artificial intelligence capabilities in a military context.

Contract Dispute and Legal Stakes

The Pentagon designated Anthropic a supply chain risk in early March 2026, marking the first public use of this procurement label against a U.S. company. This designation bars contractors from using Anthropic’s Claude AI system on Defense Department work but allows other federal agencies to continue using it. The Pentagon is expected to keep using Anthropic products for about six months while the courts resolve the dispute. The California federal court’s narrower injunction exempts non-Defense agencies from canceling existing Anthropic contracts.

The dispute began during failed negotiations over a $200 million Pentagon contract. Anthropic rejected contract language that would allow use of Claude “for all lawful purposes,” citing concerns about enabling mass surveillance of U.S. citizens and potential deployment in autonomous weapons systems. In February 2026, CEO Dario Amodei informed Defense Secretary Pete Hegseth that Anthropic would not consent to such unrestricted use.

Anthropic has filed two lawsuits against the Defense Department. The D.C. case involves a procurement statute aimed at protecting military systems from sabotage or infiltration. If broadly applied, this statute could trigger an interagency review and a government-wide blacklist. A separate California lawsuit relies on a narrower statute and led a federal judge to grant a preliminary injunction on March 26, 2026, limiting exclusion to military information systems contracts.

These conflicting rulings have created a multi-jurisdictional legal battle that could determine whether the Pentagon’s supply chain risk designation expands into a government-wide ban. The D.C. case’s expedited schedule and the statute’s broad reach raise the possibility of a single national standard, even as the California court narrows the designation’s immediate scope.

The Pentagon’s unprecedented use of the supply chain risk label gives Defense officials a new tool to restrict which AI models contractors may use. A broad application of the D.C. statute could reshape procurement decisions across federal agencies. Meanwhile, the split rulings leave contractors and civilian agencies operating under different legal conditions during the ongoing litigation.

Anthropic executives have warned the designation could cost the company billions in lost business and damage its reputation. The company remains confident the courts will ultimately rule the supply chain designations unlawful.

Defense officials maintain the label reflects a contract-compliance determination—that Anthropic declined terms the Pentagon sought—rather than retaliation for the company’s safety concerns. The acting attorney general has emphasized that military leaders require full access to Anthropic’s models to safeguard operations.