Federal Judge Calls Pentagon's Anthropic Blacklist a 'Spectacular Overreach'

A three-judge panel at the U.S. Court of Appeals for the D.C. Circuit heard oral arguments Tuesday in Anthropic's lawsuit against the Department of Defense, and the bench gave the government little comfort. Judge Karen LeCraft Henderson, one of the panel's senior members, called the Pentagon's decision to designate Anthropic a supply-chain risk a "spectacular overreach" and said she saw no evidentiary support for the label that has effectively barred defense contractors from purchasing Claude AI services since February.

The hearing was the first full airing of substantive arguments in the case, which began in March when Anthropic sued Defense Secretary Pete Hegseth in federal court. At issue is the DOD's use of a narrow procurement authority — supply-chain risk management under 10 U.S.C. § 3252 — to exclude Anthropic after the company refused to remove contractual protections prohibiting military customers from deploying Claude in mass domestic surveillance systems or fully autonomous weapons.

The Bench's Skepticism

Judge Henderson did not soften her view of the government's position. "I don't see that the department has in any way supported its determination that there is a supply chain risk with Anthropic," she told Justice Department lawyer Sharon Swingle during argument, "much less a significant supply chain risk." She pressed Swingle on whether the statutory language, written to address hardware and software procurement vulnerabilities from foreign suppliers, could plausibly reach a contract dispute with a domestic AI company over which use cases it would permit.

The exchange exposed a central weakness in the government's case: the supply-chain risk statute was designed to address foreign infiltration of defense procurement, not to give the Pentagon leverage in commercial negotiations. Henderson's framing — "spectacular overreach" — signals that at least one member of the panel views the DOD's application of the law as well outside what Congress intended.

"We're simply attempting to make sure that the department is not misusing, in our view, a narrow supply chain risk designation to gain leverage in a contract dispute — to retaliate against Anthropic for its perceived disagreement with the department." Anthropic's counsel, oral argument before the D.C. Circuit, May 19, 2026

The Government's Defense

Swingle argued that trust is the real issue, telling the panel there is a "very real prospect of new red lines" and an "increasingly hostile posture" in the company's negotiations with DOD customers. She suggested the department was entitled to weigh Anthropic's willingness to impose novel contractual restrictions when assessing supply reliability. Whether those are the kinds of risks the statute covers is precisely the question the court must answer.

Judges Gregory Katsas and Neomi Rao, the two other panel members, asked pointed questions about jurisdictional issues, including whether the court even has authority to review a supply-chain risk designation before it has worked through an administrative appeals process. The government has previously argued the designation is largely unreviewable, a claim that drew visible skepticism from the bench at earlier stages of the case. Anthropic's lawyers countered that the effective ban on Claude purchases by military contractors causes immediate commercial harm requiring court intervention now, not years from now through a separate procurement-review track.

What This Means for Anthropic

The hearing does not end the case. The panel will take the matter under advisement and issue a written opinion, which could arrive weeks or months from now. A ruling in Anthropic's favor at the circuit level could block the supply-chain designation from taking effect while the underlying dispute is resolved, or it could vacate the designation entirely. A ruling for the government would leave the blacklist standing, though Anthropic could seek Supreme Court review.

The broader commercial stakes are significant. The supply-chain designation has chilled purchasing by defense contractors who rely on government clearances, according to Anthropic. Several defense-adjacent customers have paused Claude deployments pending resolution, and the company has argued the designation is a form of retaliation for CEO Dario Amodei's refusal to modify the company's AI use-of-force policies. Amodei addressed that history in detail at the Council on Foreign Relations in May, framing the red lines as a matter of values rather than negotiating posture.

For the AI industry, the case is a test of how far existing procurement law can be stretched to govern what AI companies are allowed to refuse. If the court allows the supply-chain designation to stand, it effectively gives the executive branch a powerful tool to pressure AI labs on policy questions that go well beyond hardware sourcing. The origins of that dispute trace to January negotiations that collapsed when DOD insisted on removing Anthropic's ethical-use clauses entirely. Whatever the D.C. Circuit decides, the written opinion will become a reference point for every AI company weighing how to draft government contracts going forward.