Anthropic sues U.S. Department of Defense over supply-chain risk label

The U.S. Department of Defense designated artificial intelligence company Anthropic as a supply-chain risk—a regulatory designation used by the department to restrict business with specific entities. This designation occurred last month after Anthropic refused to make concessions on how its AI could be used by the military. Specifically, the company refused to allow its technology to be used for mass surveillance of Americans or targeting and firing decisions of lethal autonomous weapons without human intervention. The supply-chain risk label effectively bars Anthropic from working with any company that also does business with the U.S. government, as it requires those partners to certify they do not use the designated company’s products.

In response, Anthropic is suing the Department of Defense. The company argues that the designation is infringing on its First Amendment rights and punishing the company based on ideological grounds. Conversely, the Defense Department maintains that Anthropic’s refusal to allow all lawful military uses of its technology was a business decision, not protected speech, and that the designation was a straightforward national security decision rather than a punishment. The legal dispute is heading to court in San Francisco. On Tuesday, District Judge Rita Lin will preside over a hearing to decide whether to grant Anthropic a preliminary injunction—a court order to preserve the status quo while the lawsuit is litigated. To support its case, Anthropic submitted two declarations to the court last week, claiming the government’s arguments rely on technical misunderstandings and raise concerns that were not brought up during negotiations.

The Pentagon’s action has drawn sharp criticism from U.S. Senator Elizabeth Warren, who equated the decision to bar Anthropic with retaliation. In a letter to Defense Secretary Pete Hegseth, Warren wrote: “I am particularly concerned that the DoD is trying to strong-arm American companies into providing the Department with the tools to spy on American citizens and deploy fully autonomous weapons without adequate safeguards.” Warren also sent a letter to OpenAI CEO Sam Altman, requesting details about OpenAI’s agreement with the Department of Defense, which was established one day after the Pentagon blacklisted Anthropic. Meanwhile, other technology companies, including OpenAI, Google, and Microsoft, have filed amicus briefs—legal documents filed by non-parties to provide arguments—in support of Anthropic’s challenge against the designation, which is typically applied to foreign adversaries rather than American firms.

Why it matters

This case highlights the growing friction between private AI labs and the U.S. military over the ethical boundaries of deploying AI in defense and surveillance contexts. The outcome could define how much control technology developers retain over the military applications of their software.