A federal appeals court in Washington, D.C. on Friday upheld the Pentagon's designation of Anthropic as a supply chain risk, keeping the company's Claude models blocked from use by the U.S. military and its contractors. The 2-1 decision from the D.C. Circuit is a setback in Anthropic's months-long fight with the Trump administration over the Department of Defense's March 2026 blacklisting.
Circuit Judges Gregory Katsas and Neomi Rao rejected Anthropic's argument that the ban on Claude was arbitrary, unauthorized and unconstitutional. Circuit Judge Karen LeCraft Henderson dissented. The designation, first imposed in March, followed negotiations between Anthropic and the DOD over military use of Claude that broke down.
Writing for the majority, Katsas said the Pentagon's rationale cleared the statutory bar.
Key facts
- 01The D.C. Circuit upheld the Pentagon's supply chain risk designation of Anthropic in a 2-1 decision on Friday, September 25, 2026.
- 02The Department of Defense labeled Anthropic a supply chain risk in March 2026, blocking U.S. military and defense contractor use of Claude.
- 03Circuit Judges Gregory Katsas and Neomi Rao formed the majority; Circuit Judge Karen LeCraft Henderson dissented.
- 04A San Francisco federal judge ruled last month that a parallel DOD designation of Anthropic was illegal, splitting the two challenges.
- 05Anthropic said it is considering all options, including further review.
The practical effect of the designation is narrow but consequential. It prevents the U.S. military from deploying Claude across DOD information systems and prohibits defense contractors from using the models in any work performed for the agency. That closes off a procurement channel that rivals with cleared model offerings will move to fill.
Anthropic pursued the fight on two fronts because the DOD leaned on two distinct designations to justify the supply chain risk action, forcing parallel litigation in San Francisco and Washington. Last month, a San Francisco federal judge ruled that one of those designations was illegal. Friday's D.C. Circuit decision upheld the second, leaving Anthropic with a split record and the Pentagon's blacklisting intact.
An Anthropic spokesperson said the company disagreed with the ruling and pointed to the San Francisco decision as evidence the underlying government position is contested.
Anthropic has not detailed its next legal step, though the statement leaves open a request for en banc review by the full D.C. Circuit or a petition to the Supreme Court. The split between the two federal courts on parallel designations covering the same company is the kind of divergence that can accelerate higher review.
The counterweight for Anthropic is that the DOD is one customer among many. Commercial and enterprise revenue, the company's Akamai cloud deal, and international deployments are unaffected by Friday's ruling. But defense and intelligence spending on frontier models is a growing line item, and being locked out of it while competitors are not is a strategic gap that compounds over multi-year procurement cycles.
The ruling also sharpens a question the AI industry has largely been able to defer: what does it take for a frontier lab to be trusted inside classified U.S. systems, and who decides. The D.C. Circuit's answer on Friday is that the Pentagon's judgment gets substantial deference under the relevant statute, even when a sister court has reached the opposite conclusion on a parallel designation. For every model provider negotiating with the DOD, that deference is now the baseline. Anthropic's appeal, if it comes, will test whether it holds.
Working on something we should cover, or seeing a story we missed? Send leads, documents, or feedback to hello@aichatdaily.com. For sensitive tips, see our secure tips page for Signal and PGP options.
Spotted an error? Email hello@aichatdaily.com with the URL and the issue, or read our full corrections policy.
