Anthropic's Removal from US Military Supply Chain Upheld by Appeals Court

A U.S. federal appeals court on Friday upheld the Department of War's decision to remove artificial intelligence developer Anthropic from its supply chain, a significant legal victory for the government in a dispute over the military's use of frontier AI models.

Court Backs Government in AI Dispute

The U.S. Court of Appeals for the District of Columbia Circuit, in a 2–1 decision, affirmed the government's use of the Federal Acquisition Supply Chain Security Act (FASCSA) to exclude a domestic technology firm. The court found the Department had reasonable grounds to classify the safety controls in Anthropic's flagship model, Claude, as a national security risk.

Writing for the majority, Circuit Judge Gregory Katsas stated that Anthropic's contractual restrictions could disrupt or deny system functionality, falling under the statute's broad definition of a supply chain risk. The court also dismissed Anthropic's constitutional challenges, finding that the government's process was adequate given the circumstances.

Contractual Controls at Heart of Conflict

The dispute centers on Anthropic's refusal to grant the military an unrestricted "all lawful uses" contract term for its Claude AI. The company sought to maintain prohibitions on specific applications, including:

Ad

• Lethal autonomous warfare
• Mass domestic surveillance

According to the source, tensions escalated after an incident during a military operation in Venezuela in January, where Anthropic questioned if its software's deployment violated its terms of service. This raised concerns among defense officials about the reliability of mission-critical AI during active operations.

Broader Implications and Market Impact

Circuit Judge Karen LeCraft Henderson issued a dissent, arguing the majority's broad interpretation of the statute transforms it into a coercive tool to strip vendors of contractual rights. The ruling's immediate practical effect is limited by separate, ongoing litigation in a California federal court, which has issued an order lifting the government's ban.

For investors, the case highlights a growing divergence between the commercial and defense sectors for AI companies. While the Pentagon has begun shifting contracts to rival OpenAI, Anthropic's enterprise valuation continues to rise, driven by strong private sector demand. This suggests its fraught relationship with the military has not yet hindered its overall commercial trajectory.