Appeals court sides with Pentagon, keeps Anthropic blacklisted

A federal appeals court in D.C. ruled 2-1 that the Pentagon can keep Anthropic on its national-security blacklist. The March supply chain risk designation stands, so Claude stays off parts of the Defense Department's systems. Anthropic says the label has already cost it billions in lost business, and the ruling lands as the company heads toward an expected IPO.

The dispute goes back to a $200M Pentagon contract signed in July 2025. Talks over putting Claude on the DOD's GenAI.mil platform fell apart when the military asked for use "for all lawful purposes." Anthropic wanted a guarantee that its models would not power fully autonomous weapons or domestic mass surveillance. The majority called the designation reasonable and rejected Anthropic's claim that it was retaliation for its safety views.

The courts now disagree with each other. Last month a San Francisco federal judge ruled that a parallel DOD designation was unlawful and found First Amendment retaliation. Anthropic says it "respectfully disagrees" and is weighing a request for en banc review by the full appeals court.

Why it matters: a court has now said the government can punish an AI lab for the limits it puts on its own models, which puts pressure on every lab chasing defense contracts.