Appeals court upholds Pentagon's designation of Anthropic as a supply chain risk

Translated by a model; the original is in Russian.

· Artificial intelligence
A federal appeals court in Washington backed the Defense Department's blacklisting of Anthropic, which bars the military from using its Claude models.

cover.svg

A federal appeals court in Washington, D.C., on Friday upheld the Pentagon's decision to blacklist Anthropic as a supply chain risk, dealing another blow to the AI company in its months-long fight with the Trump administration.

The three-judge panel ruled 2-1, rejecting Anthropic's argument that the Defense Department's ban on its Claude models was arbitrary, unauthorized and unconstitutional. The majority consisted of Judges Gregory Katsas and Neomi Rao, both appointed by Donald Trump; Judge Karen Henderson, appointed by George H.W. Bush, dissented.

How the dispute began

In March the DOD labeled Anthropic a supply chain risk, effectively accusing it of threatening national security. The designation bars the military from using the company's models and blocks defense contractors from using them in agency work. The action rested on two separate justifications, which had to be litigated in two different courts. Last month a federal judge in San Francisco found one of them unlawful; Friday's ruling in Washington upheld the second.

The conflict erupted after negotiations collapsed. In July 2025 Anthropic signed a $200 million contract with the Pentagon, but by fall talks over deploying Claude on the DOD's GenAI.mil platform broke down. The department demanded unfettered access to the models for all lawful purposes, while the company insisted on assurances that its technology would not be used for fully autonomous weapons or domestic mass surveillance. Defense Secretary Pete Hegseth accused Anthropic of trying to "seize veto power over the operational decisions of the United States military."

The court's reasoning

Katsas cited Hegseth's "deeply sobering" concern that overly constrained AI models could shut down unexpectedly or be subject to manipulation. Anthropic disputed those claims, but the court held that final authority rests with the president and the secretary, who must weigh the competing risks.

The panel delayed its decision from taking immediate effect to give Anthropic time to seek a rehearing — by the same panel or by the full D.C. Circuit. The company could also petition the Supreme Court. "We respectfully disagree with the court's decision... and are considering all options," an Anthropic spokesperson said.


Source: Hacker News

Comments

Sign in, to comment.

  • No comments yet.