28 August 2026

Judge blocks Pentagon's blacklisting of AI company Anthropic

• A federal judge ruled the Pentagon illegally designated Anthropic a supply chain risk, barring defense contractors from using its Claude AI models, based on retaliation for the company's public criticism of military AI policy.
• The dispute started over a contract disagreement: Anthropic refused to remove restrictions on using Claude for autonomous weapons and mass surveillance, while the Pentagon demanded unrestricted military access to the technology.
• The judge found the Pentagon's legal justifications baseless, including its claim that Anthropic could secretly disable its models once deployed in military systems, which officials later conceded was false.
• A separate lawsuit in Washington, D.C. remains ongoing, so Anthropic technically stays blacklisted until that case resolves, though the Pentagon is expected to appeal the California ruling.

How it was covered

The AlgorithmMIT Technology Review

A court ruled that the Pentagon's designation of Anthropic violated First Amendment rights and was based on retaliation. Anthropic still faces a separate case in Washington, DC.

Reported by CNBC, Wired, SiliconANGLE, The Guardian, Ars Technica, The New York Times