Court Upholds Pentagon Anthropic Ban

A US appeals court has upheld the Pentagon's decision to exclude artificial intelligence company Anthropic from the Defense Department's supply chain, deepening a dispute over the company's refusal to remove safeguards from its Claude AI model.

The court issued its ruling on Friday, allowing a Pentagon designation that prevents the department and its contractors from using Anthropic's technology to remain in effect.

The dispute began in February, when Anthropic declined a demand from then-Defense Secretary Pete Hegseth to remove restrictions governing the use of Claude for fully autonomous lethal weapons and large-scale domestic surveillance of Americans.

The Pentagon subsequently designated Anthropic a "supply chain risk" in March. The decision resulted in the cancellation of military contracts and prohibited other Defense Department contractors from using the company's technology.

Anthropic has argued that the designation has cost the company billions of dollars in potential business and harmed its reputation as it prepares for a closely watched initial public offering.

The action against Anthropic is considered an unusual use of the Pentagon's supply-chain authority against a US technology company. The dispute has placed the AI developer at odds with the Trump administration despite Anthropic's prominent position in the rapidly expanding artificial intelligence industry.

Anthropic chief executive Dario Amodei has repeatedly warned about potential risks associated with increasingly powerful AI systems. President Donald Trump has dismissed such warnings, describing them as a hoax.

Friday's appeals court ruling also creates a legal conflict with a separate decision issued by a federal court in California in August. That court found that a broader ban imposed on Anthropic by the Trump administration was unlawful.

The Pentagon's action upheld by the appeals court was based on federal public procurement regulations and remains in effect.

Anthropic said it disagreed with the latest ruling and indicated that it was considering further legal action.

"We respectfully disagree with the court's decision," an Anthropic spokesperson said. The company said another federal court had already found the government's parallel designation unlawful and that it remained confident in its legal position.

Claude has reportedly been used by the US military across a range of classified and sensitive systems. The technology was also reportedly involved in military operations, including a January operation that resulted in the capture and removal of Venezuelan leader Nicolas Maduro.

The legal dispute highlights the growing tension between government demands for access to advanced AI systems and technology companies' restrictions on how their models can be deployed, particularly in military and domestic surveillance applications.