A federal appeals court in Washington, D.C., has upheld the Pentagon’s 2–1 decision to designate AI company Anthropic as a “supply chain risk,” marking a significant development in the company’s legal dispute with the US government. The ruling rejected Anthropic’s argument that the Department of Defense’s restrictions on its Claude AI models were arbitrary, unauthorized, and unconstitutional.
The majority found that the Pentagon had sufficient grounds to consider continuing use of Anthropic’s technology a potential national-security risk. The designation, imposed in March 2026, effectively bars the US military from using Anthropic’s Claude models and restricts defense contractors from employing the company’s technology in Department of Defense projects. Anthropic challenged the government’s action in court, seeking to overturn the designation and oppose the restrictions.
The dispute arose after Anthropic signed a $200 million contract with the Pentagon in July 2025 and began negotiations to deploy Claude through the military’s GenAI.mil platform. While the Pentagon sought access to Anthropic’s models for all lawful military uses, the company requested assurances its technology would not be used for fully autonomous weapons or domestic mass surveillance. The parties failed to reach an agreement, leading to a breakdown in their relationship.
In its ruling, the appeals court majority endorsed the Pentagon’s authority to assess potential risks from using Anthropic’s AI systems in defense operations. Judge Gregory Katsas stated that decisions balancing competing national-security risks rest with the president and the defense secretary. Judge Karen LeCraft Henderson dissented.
This ruling is part of a broader legal battle involving two separate government designations against Anthropic. A federal judge in San Francisco previously ruled in August that one designation was unlawful. The D.C. appeals court’s decision upheld the separate designation challenged in its jurisdiction, creating differing outcomes between the courts.
Anthropic said it respectfully disagrees with the appeals court’s decision and referenced the earlier San Francisco ruling. The company remains confident in its legal position and is considering further options. The appeals court has delayed immediate implementation of its decision, allowing Anthropic time to seek a rehearing, request full-court review, or petition the US Supreme Court.
The case underscores the growing importance of AI technology in national security and defense and the disputes over governmental and military use of advanced AI systems. It also imposes additional restrictions on Anthropic’s participation in Pentagon projects while the company pursues further legal remedies.
