Anthropic has asked the full US Court of Appeals for the District of Columbia Circuit to rehear its challenge to the Pentagon’s decision to exclude its Claude AI models from military systems, CNBC reported on 13 October 2026, in the latest twist in a legal fight that has split the courts and reshaped the military’s AI supplier list.
The AI company filed the rehearing petition on 13 October 2026, asking the full appeals court — or at minimum the original three-judge panel — to revisit a divided 2-1 ruling issued on September 25, 2026, that upheld the Pentagon’s designation of Anthropic as a national-security supply-chain risk. The petition was reported first by CNBC, according to the reports.
The case turns on a federal supply-chain law written to target hostile foreign actors. Anthropic argues the statute cannot be used against a US-based AI lab that refused to let its models be used for mass surveillance or autonomous weapons. The Pentagon says Claude’s continued integration into its systems, including through contractors, is a statutory supply-chain risk.
Two courts, two different answers
The dispute has produced contradictory outcomes in different courts, which is why the rehearing request matters. On August 27, 2026, US District Judge Rita Lin in the Northern District of California set aside a separate Pentagon designation of Anthropic under a different law, calling the decision illegal and arbitrary. But the D.C. Circuit panel ruled on a distinct exclusion under 41 U.S.C. Section 4713, part of the Federal Acquisition Supply Chain Security Act, and denied Anthropic’s petitions for review.
Circuit Judges Gregory Katsas and Neomi Rao formed the majority, with Katsas writing the opinion, while Circuit Judge Karen Lecraft Henderson dissented, according to reports. The split is exactly what Anthropic is leaning on in its rehearing request, since a 2-1 decision leaves room for the full court to see the statutory question differently.
The underlying dispute dates to February 2026, when Defence Secretary Pete Hegseth summoned Anthropic chief executive Dario Amodei to the Pentagon. Anthropic said it turned down a request for unrestricted access to its models, arguing that modern models are insufficiently reliable for autonomous lethal weapons and that mass domestic surveillance conflicts with democratic principles. Hegseth then moved to designate the company a supply-chain risk, and President Trump ordered all federal agencies to stop using its technology.
OpenAI and rivals fill the gap
While Anthropic fights in court, its rivals have moved in. The Pentagon announced agreements with seven tech firms, including OpenAI, Alphabet, Nvidia, SpaceX, Microsoft, Amazon and Reflection, to use their AI tools on classified military networks. OpenAI signed its deal shortly after Anthropic was labelled a threat, though the agreement was later adjusted to prevent its tools being used for domestic surveillance of American citizens, following public pushback.
Anthropic sued the Trump administration in March, accusing the government of an unlawful campaign of retaliation. The Department of War — the renamed Department of Defense — halted use of Anthropic AI after the appellate ruling in September, according to reports. The exclusion stays in effect unless the full D.C. Circuit agrees to rehear the case and reverses it, or Anthropic takes the fight to the Supreme Court. For now, procurement teams are treating the Pentagon exclusion as active, and the rehearing decision is the next concrete step to watch.
Sources: Tech Insider, NeoTeo, EverMX, AIN.UA, Forbes; CNBC, 13 Oct 2026
