The U.S. Court of Appeals for the District of Columbia Circuit ruled 2-1 on September 25, 2026, to uphold the Pentagon's designation of Anthropic as a "supply chain risk" under the Federal Acquisition Supply Chain Security Act. The decision, authored by Circuit Judge Gregory Katsas, rejected Anthropic's challenge on statutory and constitutional grounds, finding the Department of War had "ample support" for its determination that continuing to integrate Claude into military information systems posed a national security risk.
The ruling stems from a dispute that escalated after Anthropic refused to relax contractual prohibitions on the use of Claude for lethal autonomous warfare and domestic surveillance. On March 3, 2026, Secretary of Defense Pete Hegseth announced the designation, effectively barring Claude from all Department of War systems and networks. The move was unusual: supply chain risk designations under the 2018 statute have historically been reserved for foreign adversaries, not domestic AI companies.
In its majority opinion, the court found that Anthropic's own admission that it encodes restrictions into Claude to prevent certain uses was central to the government's case. "On more than one occasion, these restrictions have stopped Claude from performing tasks requested by government users," Katsas wrote. The court also cited a recent dispute over whether contractual prohibitions barred Claude's use in an ongoing overseas military operation, leaving the department uncertain whether the model "would perform as needed and intended."
Anthropic argued the designation violated its due process and First Amendment rights. The court rejected both claims. On due process, the panel found the department promptly notified Anthropic and gave it a fair opportunity to contest the exclusion. On the First Amendment, Katsas wrote that the exclusion was based on Anthropic's refusal to comply with an essential contractual obligation, not on its constitutionally protected advocacy for AI safety regulation.
Judge Karen Henderson dissented, arguing the panel should have applied a narrower definition of "supply chain risk" under the statute. She warned the decision could empower the secretary to coerce any AI contractor into compliance. "Suppose the secretary tells Anthropic's presumed replacement to change its AI-use policies to permit any 'functions that the department deems necessary' or it will share the same fate," Henderson wrote. "According to today's decision, that contractor will have a choice: Agree to the secretary's demands or risk being designated a national security threat."
The implications for Anthropic are significant. The company reported annualized revenue of $65 billion and has an IPO expected in the coming weeks. The blacklisting has canceled military contracts and, according to Anthropic, caused billions of dollars in losses while damaging its reputation ahead of the public offering. A spokesperson said the company "respectfully disagrees" with the ruling and "remains confident in its position and is considering all options, including further review."
The ruling deepens a split in federal courts. In August, a Northern California District Judge found the government's parallel designation unlawful, ruling it violated the First Amendment. The D.C. Circuit's decision on Friday directly contradicts that finding, raising the prospect of an eventual Supreme Court battle over the scope of executive power to regulate AI systems used in national defense.
Sources
- [Court opinion (PDF)](media.cadc.uscourts.gov) - United States Court of Appeals for the District of Columbia Circuit, No. 26-1049, decided September 25, 2026
- [DC Circuit finds Pentagon justified in labeling Anthropic supply chain risk](courthousenews.com) - Courthouse News Service
- [US court upholds Pentagon's blacklisting of Anthropic](aljazeera.com) - Al Jazeera
- [D.C. appeals court sides with Pentagon on blacklisting Anthropic](thehill.com) - The Hill
