US appeals court upholds Pentagon supply chain ban against Anthropic over AI safeguards

A federal appeals court on Friday upheld the Pentagon's authority to exclude artificial intelligence developer Anthropic from its military supply chain after the startup refused to eliminate user safety restrictions on its Claude AI model.

Why did the Pentagon designate Anthropic as a defense supply chain risk?

The dispute originated in February 2026 when Defense Secretary Pete Hegseth demanded Anthropic lift contractual safeguards prohibiting the use of Claude for fully autonomous lethal weapons and mass domestic surveillance of Americans. When Anthropic refused, the Pentagon formally designated the firm a supply chain risk in March, canceling military contracts and prohibiting defense contractors from utilizing its technology.

Tensions escalated after an Anthropic executive questioned defense contractor Palantir over its deployment of Claude during the January 3 military operation that captured Venezuelan President Nicolás Maduro. In a 2–1 decision, US Circuit Judge Gregory Katsas wrote for the DC Circuit majority that the Department of Defense maintained sufficient statutory authority under supply chain security laws to address operational risks, noting that embedded corporate restrictions created uncertainty over whether the software could refuse authorized commands or disrupt active military missions.

How does the ruling impact military AI contracts and Anthropic's legal standing?

Following Anthropic's procurement designation, the Pentagon turned to competing AI developers, expanding contracts with OpenAI, xAI, Microsoft, and Google to deploy models across sensitive defense networks. Under Secretary of Defense Emil Michael welcomed the appellate ruling, stating that military command structures must operate without private corporate intervention.

The DC Circuit decision conflicts with a separate decision issued last month by a federal district court in California, which found a parallel executive branch ban against Anthropic unlawful. An Anthropic spokesperson stated that the company respectfully disagrees with Friday's ruling, pointing to its favorable California judgment while considering further legal review—including a potential appeal to the US Supreme Court—as it prepares for an initial public offering.