Federal appeals court upholds Pentagon's risk label on Anthropic's Claude

A federal appeals court in Washington, DC, ruled Friday that the Pentagon can keep its supply-chain risk label on Anthropic, rejecting the company’s bid to have the designation thrown out. The 2-1 decision keeps Claude, Anthropic’s AI model, blocked from military and other federal government systems. The majority opinion said the Department of Defense had “ample support for its conclusion that the continued integration of Claude into the department’s information systems… presented a statutorily covered national-security risk.” The judges pointed to Anthropic’s own admission that it builds restrictions into Claude that stop the model from carrying out tasks the company objects to.

Anthropic spokesperson Danielle Cohen said the company remains confident in its position and is weighing its options, which could include asking a larger panel of the DC Circuit Court of Appeals to rehear the case or petitioning the Supreme Court.

The Pentagon issued the designation earlier this year under two separate supply-chain laws, ordering Claude removed from military systems and other parts of the federal government by this month. Anthropic executives had said they would not let the government use the company’s current models to support autonomous weapons or domestic surveillance. Defense Secretary Pete Hegseth called that stance a serious national security risk.

Because the two laws fall under different courts, Anthropic had to fight the designations separately. A federal judge in San Francisco struck down one of the labels in March and reaffirmed that ruling last month. Friday’s decision means the second designation, the one reviewed by the DC Circuit, stays in place indefinitely. Both cases could still take years to work through further appeals.

What the ruling rejected

The majority also dismissed Anthropic’s argument that the designation violated its due process and free speech rights. The judges wrote that the Pentagon “excluded Anthropic from its supply chain based on the company’s refusal to assent to a contract term that the Department deemed essential, not based on the company’s support for greater governmental regulation of AI technology.” They described the dispute as an ordinary contract disagreement rather than retaliation, and rejected Anthropic’s separate claim that the government had acted beyond what the supply-chain law allows. The outcome was not entirely a surprise: the same three-judge panel had declined in April to temporarily block the designation, finding Anthropic had not met the “stringent requirements” for emergency relief, and judges appeared divided on the underlying question before settling on a 2-1 result.

Anthropic said soon after the designations took effect that it lost revenue as customers grew wary of doing business with a company the government had flagged. It has not given an updated figure since, and has continued to report growing sales overall as it moves toward a planned initial public offering later this year. The Pentagon, for its part, has not detailed how far along it is in replacing Claude with alternatives such as SpaceX’s Grok, Google’s Gemini, or OpenAI’s GPT models. Some employees at Google and OpenAI have objected to their companies taking on military work that Anthropic turned down, though both companies have said supporting the US government remains a priority. Whether Anthropic pursues a further appeal, and how that affects its IPO timeline, is the next thing to watch.