Court Voids Pentagon's Blacklisting of Anthropic
Judge Rita Lin found the supply chain risk label unconstitutional, ruling that the Pentagon punished Anthropic for criticizing US AI policy.
Illustration: an empty courthouse rotunda in morning light, with server cabinets hinted in the background.
A federal district court in California ruled on August 28, 2026 that the Pentagon's supply chain risk designation against Anthropic was unlawful, because it penalized the company for publicly criticizing the administration's AI policy.
At a glance
- Ruling issued August 28, 2026 by Judge Rita Lin, US District Court for the Northern District of California.
- The designation labeled Anthropic a supply chain risk and barred defense contractors from using Claude.
- The Defense Department conceded during the case that its claim of covert access to fielded systems was untrue.
- Anthropic filed suit in March 2026, while the Pentagon was using Claude internally.
- The ban stays formally in place until a second proceeding in Washington is resolved.
A federal judge has struck down the Pentagon's ban on Anthropic's models, ruling on August 28, 2026 that labeling the company a supply chain risk was retaliation for speech. Judge Rita Lin of the US District Court for the Northern District of California called the ban illegal and baseless.
What the court found
Two government justifications collapsed under scrutiny. The first was an allegation that Anthropic held some form of backdoor access to systems already deployed inside government. The Defense Department conceded during the proceedings that the claim was untrue, and the court recorded that Anthropic undisputedly holds no such access.
The second justification rested openly on the company's public positions, including an essay by chief executive Dario Amodei published in January. Per SiliconANGLE, Lin wrote that the Constitution does not permit sweeping penalties imposed principally because of criticism.
How the ban started
The Pentagon approached Anthropic in the fall of 2025 about putting Claude on the GenAI.mil platform. Talks broke down over usage terms. Anthropic would not drop contract language that bars customers from running mass surveillance of Americans or building autonomous weapons, and the department wanted unrestricted access.
By late 2025 federal agencies were barred from using Claude, and Defense Secretary Pete Hegseth added the supply chain risk designation, which extended the prohibition to defense contractors. Anthropic sued in March 2026 — at a moment when, according to SiliconANGLE, the Pentagon had begun using Claude internally.
Why the ban has not lifted yet
The California decision does not end the dispute. A second proceeding in Washington remains open, and the designation stands until it concludes. SiliconANGLE identifies that forum as the US Court of Appeals for the D.C. Circuit; The Decoder describes it more generally as a pending Washington case. Either way, Anthropic remains technically blacklisted for now.
What is at stake for Anthropic
The company said it wants to keep working with the government on national security uses of AI, without giving up its own limits on how the models may be used. The Decoder notes the timing against a planned fall IPO: a live blacklisting weighs more heavily while federal agencies are becoming among the largest buyers of AI.
One further report could not be checked for this article. The Verge's coverage of the same ruling was not retrievable, so any detail unique to it — exact language from the opinion, or contract values — is not reflected here.
FAQ
Why did the Pentagon blacklist Anthropic?
After negotiations over military use of Claude collapsed, the Defense Department designated the company a supply chain risk, citing among other things its public criticism of government AI policy. The court held that reasoning impermissible.
Can the US military use Claude again now?
Not automatically. The California ruling went Anthropic's way, but a second proceeding in Washington is still pending, and the designation formally remains in force until then.
What was the Anthropic Pentagon dispute actually about?
Usage limits. Anthropic refused to remove contractual bans on mass surveillance of Americans and autonomous weapons development, while the department sought unrestricted access to the models.
