D.C. Circuit Upholds Pentagon's Blacklisting of Anthropic in 2-1 Ruling, Splitting From an Earlier Court Win for the Company

Overview

A federal appeals court ruled Friday that the Pentagon can continue to treat Anthropic as a “supply chain risk,” upholding a designation that bars the U.S. military and its contractors from using the company’s Claude models. The 2-1 decision from a three-judge panel of the U.S. Court of Appeals for the District of Columbia Circuit found the Department of Defense had authority to restrict Anthropic’s access under federal procurement law, according to Ars Technica. The ruling is a setback for Anthropic that comes just weeks after the company won a separate court fight over a different Pentagon designation, according to CNN.

What We Know

Judges Gregory Katsas and Neomi Rao, both appointed by President Trump, formed the majority, while Judge Karen LeCraft Henderson, appointed by President George H.W. Bush, dissented, according to CNBC. The panel found the Trump administration had authority under 41 U.S.C. § 4713 to restrict Anthropic’s technology access for federal agencies and defense contractors, according to Ars Technica. The majority wrote that “the case raises profoundly difficult questions about the appropriate military uses of an almost unimaginably powerful new technology,” weighing the risk of “overly constrained AI models shutting down unexpectedly and thus causing important military operations to fail” against “unconstrained AI models hallucinating inappropriate targets for lethal military force,” per Ars Technica. The court concluded that “the Secretary did not transgress any limits on his authority under the Supply Chain Security Act or the Constitution,” and that Anthropic’s own restrictions on Claude’s military use “qualifies as a ‘supply chain risk’ within the meaning of section 4713,” the outlet reported.

In dissent, Judge Henderson argued the statute, “enacted…in response to calls from the US intelligence community,” was meant to target hostile actors rather than “a contractor’s honest and upfront enforcement of restrictions on a covered article’s use disfavored by the government,” according to Ars Technica.

The Pentagon designated Anthropic a supply chain risk in March 2026, a label that “prevents the U.S. military from using Claude models and blocks defense contractors from using them in work with the agency,” CNBC reported. The designation was, according to CNN, “traditionally reserved for foreign adversary-linked firms and never previously applied to an American company.” Defense Secretary Pete Hegseth had argued that “overly constrained” AI models risked “shutting down unexpectedly” and that Claude might be “subject to manipulation,” accusing Anthropic of trying to “seize veto power over the operational decisions of the United States military,” according to CNBC. Pentagon Undersecretary Emil Michael celebrated the ruling on social media, writing that “the hammer of justice has smashed @AnthropicAI arguments” and that “warfighters will sleep better knowing that no private company will insert their opinions in the chain of command,” CNN reported.

The dispute traces back to a Defense Department contract worth up to $200 million that Anthropic was awarded in July 2025 alongside Google, OpenAI and xAI, as previously reported. Anthropic refused Pentagon demands to let Claude be used for fully autonomous weapons and domestic mass surveillance of U.S. citizens, a standoff that led to its blacklisting in February, as previously reported.

An Anthropic spokesperson said, “We respectfully disagree with the court’s decision” and that the company is “considering all options, including further review,” according to CNBC. The spokesperson added, “Another federal court has already held the government’s parallel designation unlawful,” according to CNN.

What We Don’t Know

Anthropic has not said whether it will petition the full D.C. Circuit for an en banc rehearing or take the case directly to the Supreme Court — both of which remain open options, according to CNBC. The appellate panel delayed implementation of its ruling to give Anthropic time to pursue further review, the outlet reported.

Analysis

Friday’s ruling does not overturn the win Anthropic secured in August, when a federal judge found a separate, parallel Pentagon designation against the company unlawful. The two cases turn on different statutes: the district court judge, Rita Lin, ruled under 10 U.S.C. § 3252, which requires a finding of “bad motive” or adversarial intent, while the D.C. Circuit reviewed the case under 41 U.S.C. § 4713, which the appeals court said carries no such requirement, according to Ars Technica. The government has not appealed Lin’s ruling, and it still allows other federal agencies and their contractors to keep working with Anthropic, according to CNN — but the Pentagon itself remains off-limits to Claude under the newly upheld designation. The split outcome leaves Anthropic barred from military use while it pursues further appeals, at a moment when the company is preparing for what CNN described as a “potentially trillion-dollar” initial public offering that depends in part on its enterprise business.