Court upholds Pentagon’s supply-chain risk label on Anthropic’s Claude AI

5 hours ago 29

A federal appeals court in Washington, D.C. has sided with the Pentagon in one of the most closely watched legal fights in the artificial intelligence industry this year, upholding a decision that brands Anthropic a supply-chain risk and keeps its Claude models locked out of most of the federal government. The 2-1 ruling, handed down Friday, marks a real setback for the AI lab and shows just how much leverage the Trump administration now holds over which AI companies get to work with the military. The dispute over the Pentagon Anthropic supply chain designation is far from settled, though — a separate court has already gone the other way, and Anthropic says it’s weighing further legal options.

Key takeaways

A DC Circuit Court of Appeals panel voted 2-1 on Friday to uphold the Pentagon’s designation of Anthropic as a supply-chain risk.
The ruling keeps Anthropic’s Claude AI models barred from use by the U.S. military and its contractors.
Anthropic’s due process and free speech arguments were rejected; the majority said the government simply followed standard contract procedure.
A San Francisco federal judge separately struck down a related supply-chain risk label against Anthropic in March, confirming that ruling last month.
The Pentagon is looking at alternatives to Claude, including SpaceX’s Grok, Google’s Gemini, and OpenAI’s GPT models.

Federal Appeals Court Upholds Pentagon’s Supply-Chain Risk Label

The court’s majority found that the Department of Defense had solid legal grounding for its move, writing that “the continued integration of Claude into the department’s information systems, by the department or its contractors, presented a statutorily covered national-security risk.” Judges Gregory Katsas and Neomi Rao, both Trump appointees, formed the majority opinion, while Judge Karen LeCraft Henderson dissented.

Court’s Reasoning on National Security and Contract Refusal

Crucially, the judges made clear this wasn’t about punishing Anthropic for its views on AI regulation. The Pentagon “excluded Anthropic from its supply chain based on the company’s refusal to assent to a contract term that the Department deemed essential, not based on the company’s support for greater governmental regulation of AI technology,” the panel wrote. In other words, this was framed as an ordinary contract dispute rather than a policy crackdown — a distinction that matters a lot for how future supply-chain risk designations against other AI firms might hold up in court.

The majority also dismissed Anthropic’s claims that its due process and free speech rights had been violated, saying the government had followed proper procedure throughout.

What Changes Now for Claude in Military and Government Systems

In practical terms, the ruling means Claude stays off the table for the U.S. military and any federal contractors working with the Pentagon, extending a ban that was originally set to take effect this month. Earlier this year, the Department of Defense sanctioned Anthropic under two separate supply-chain risk laws, aiming to strip Claude out of military systems and other corners of the federal government.

The root of the standoff goes back to a dispute over how far the Pentagon could push Anthropic’s technology. Anthropic executives had refused to let the government deploy Claude for autonomous weapons systems or domestic surveillance, a red line the company would not cross. Secretary of Defense Pete Hegseth deemed the stance a significant national security risk.

This is where the case matters beyond Anthropic itself: the ruling effectively tells AI companies that if they want to work with the Pentagon, they may have to accept less control over how their models get used in sensitive military contexts, or risk getting shut out entirely.

Two Courts, Two Outcomes — and a Possible Appeal Ahead

What makes this fight so tangled is that the Pentagon relied on two separate supply-chain risk designations against Anthropic, and each one had to be challenged in a different court. A federal judge in San Francisco already ruled in March — and reaffirmed last month — that one of those designations was illegal. Friday’s decision in Washington upheld the other one, meaning it stays in place indefinitely for now.

That split outcome leaves Anthropic in an odd legal limbo: partially cleared in one venue, still blocked in another. Anthropic spokesperson Danielle Cohen made clear the company isn’t done fighting. “We respectfully disagree with the court’s decision,” Cohen said in a statement. “Another federal court has already held the government’s parallel designation unlawful. We remain confident in our position and are considering all options, including further review.”

Those options include asking the same three-judge panel for a rehearing, requesting a full en banc review by every judge on the DC Circuit Court of Appeals, or ultimately petitioning the Supreme Court. Given the years of appeals both cases are likely to face before final resolution, this legal battle over the Pentagon Anthropic supply chain dispute could easily stretch well into the future.

Pentagon’s Alternatives and Anthropic’s Business Path Forward

With Claude effectively locked out, the Pentagon is looking elsewhere. Reporting indicates the government wants to replace it with AI models from SpaceX’s Grok, Google’s Gemini, and OpenAI’s GPT — though the Department of Defense hasn’t shared detailed updates on how that transition is actually going. Notably, some employees at Google and OpenAI have raised ethical objections internally about their companies taking on military contracts that Anthropic itself walked away from, but both companies have pushed past those objections and continued to frame their government work as important.

For Anthropic, the fallout has a real financial dimension. The company said early on that it lost revenue as customers grew wary of doing business with a firm branded a government risk, though it hasn’t detailed exactly how deep that impact has run. Even so, Anthropic has continued to tout rising sales more broadly and is still moving toward a potential initial public offering later this year — suggesting that, for now at least, the Pentagon standoff hasn’t derailed the company’s larger commercial ambitions.

That tension — a company losing ground with one of the world’s largest institutional buyers while still chasing a public listing — is worth watching closely. It raises a broader question for the AI industry: how much government business can a company afford to walk away from on principle before it starts showing up in the numbers investors care about?

FAQ

Why did the Pentagon designate Anthropic as a supply-chain risk?

The Pentagon designated Anthropic as a supply-chain risk because it viewed the company’s refusal to allow its AI to be used for autonomous weapons or surveillance as a national security risk.

What does the court ruling mean for Anthropic’s AI models?

The ruling prevents the Pentagon and its contractors from using Anthropic’s Claude AI models in military and federal government systems.

Can Anthropic appeal the federal appeals court decision?

Indeed, Anthropic is weighing whether to challenge the decision before either a larger panel of the DC Circuit Court of Appeals or the US Supreme Court.