Anthropic's Classified Dreams Crushed: Pentagon Sticks AI Firm with Unprecedented 'Supply Chain Risk' Label That Won't Budge

The US Court of Appeals for the DC Circuit has rejected Anthropic's attempt to pause a Pentagon designation labeling the company a national security supply chain risk. A three-judge panel denied the emergency motion for a stay on Wednesday, ruling that the government's interest in controlling how it secures AI technology during active military conflict outweighs any financial or reputational harm Anthropic may suffer. Looks like Claude's clearance to peek at classified docs got rug-pulled harder than a DeFi protocol in a bear market.

This means part of the Department of Defense's official designation of Anthropic's products as a "supply-chain risk to national security" remains in place. It's a first for an American company and bars Pentagon contractors from using Anthropic's AI models. Anthropic just earned the dubious honor of being the first tech unicorn to get blacklisted harder than a mixer on chain analysis software—no one wanted this badge of honor, but here we are.

The dispute traces back to a July 2025 deal where Anthropic was set to make Claude the first large language model approved for classified networks. Those negotiations collapsed in February when the government demanded unrestricted military use of Claude. Anthropic pushed back, insisting its technology shouldn't power lethal autonomous weapons or mass domestic surveillance of Americans. Think of it as the ultimate "no means no" moment, except the Pentagon was really hoping for a "yes and" from their new AI buddy.

President Donald Trump ordered all federal agencies to stop using Anthropic products in late February, calling the company's position a "disastrous mistake trying to strong-arm the Department of War." Anthropic sued the administration in March, calling it an "unlawful campaign of retaliation." Nothing says "we're taking our toys and going home" quite like a presidential tweetstorm and a lawsuit filed faster than you can say "constitutional crisis."

A Northern District of California court initially granted a preliminary injunction against the Pentagon's designation and temporarily halted Trump's directive, calling it "Orwellian." But federal procurement law forced Anthropic to fight the battle on two separate legal tracks—constitutional grounds in California and directly at the D.C. Circuit under the specific statute authorizing the designation. Being forced to litigate on two fronts simultaneously is the legal equivalent of getting front-run by everyone while trying to execute a perfectly reasonable transaction.

The appellate ruling acknowledged Anthropic will "likely suffer some degree of irreparable harm absent a stay" but found the government's interests paramount. Acting US Attorney General Todd Blanche celebrated it as a "resounding victory for military readiness," stating that "military authority and operational control belong to the Commander-in-Chief and Department of War, not a tech company." Nothing like watching the DOJ flex on an AI startup like it's arug-pull in slow motion—except this time the token was "democracy" and the rug was pulled by the federal government.