The AI startup Anthropic has asked a federal appeals court for an emergency stay to halt a Pentagon order that designates the company as a “supply chain risk.” In a filing with the U.S. Court of Appeals for the D.C. Circuit, the company warned that the label—typically reserved for foreign adversaries—could cost it billions of dollars in lost revenue and cause “irreparable harm” to its reputation.

The Root of the Conflict The legal battle follows a breakdown in negotiations between Anthropic and the Department of Defense. The company refused to remove “ethical guardrails” on its AI model, Claude, specifically requesting that the technology not be used for mass domestic surveillance or to power fully autonomous lethal weapons. In response, Defense Secretary Pete Hegseth labeled the firm a security risk and banned its use by the Pentagon and its contractors.

Economic Fallout Anthropic’s legal team argues that the government’s actions are a form of unlawful retaliation for the company’s stance on AI safety. The impact has been immediate:

• Revenue at Risk: The company estimates the designation could slash its 2026 revenue by “hundreds of millions, or even multiple billions, of dollars.”
• Customer Anxiety: More than 100 enterprise clients have reportedly contacted Anthropic to express concern over the “supply chain risk” tag, which could theoretically affect any company doing business with the Department of Defense.
• Secondary Effects: Anthropic claims the move has already chilled its commercial business, leading some non-government clients to pause or cancel contracts out of fear of future federal scrutiny.

Industry Support The case has drawn significant attention from across the tech sector. Microsoft filed a motion in support of Anthropic, arguing that the Pentagon’s move could destabilize the entire American AI ecosystem. Additionally, a group of researchers from rivals OpenAI and Google signed a joint letter supporting Anthropic’s refusal to allow its models to be used for autonomous targeting.

The Pentagon’s Stance While the Pentagon has not formally responded to the latest request for a stay, an internal memo suggests the military is struggling to find a replacement for Claude. The memo noted that while a six-month phase-out is underway, some “mission-critical” activities may receive rare exemptions if no viable alternative to Anthropic’s technology can be found.

What’s Next? Anthropic is currently pursuing two legal tracks: the request for a stay in D.C. and a broader lawsuit in California challenging the constitutionality of the blacklist. The courts must now decide whether the executive branch has the authority to label a domestic company a “supply chain risk” based on a disagreement over software usage terms.