Three months after receiving a "supply chain risk" designation from the Department of War, Anthropic is fighting back in court.
The company filed suit challenging the designation, arguing it amounts to government retaliation for Anthropic's refusal to agree to "all lawful use" terms. Those terms would have permitted unrestricted deployment of Claude for surveillance and autonomous weapons systems - without the safety guardrails Anthropic builds into its products.
"Remove Your Safety Guardrails or Else"
The dispute centers on contract language. The Department of War wanted Anthropic to drop usage restrictions that prevent mass surveillance of Americans without judicial oversight and block autonomous weapons authorization. When Anthropic refused, the government issued a supply chain risk (SCR) designation - a label that effectively bars federal agencies and contractors from doing business with the company.
Anthropic's legal argument has two prongs: the SCR designation lacks statutory authority, and it violates First Amendment protections. The company points to public statements by government officials explicitly linking the designation to Anthropic's policy positions on AI use restrictions. In legal terms, that's "jawboning" - using government pressure to coerce a private company into changing its behavior.
When Your Competitors Back Your Lawsuit
The support list tells its own story. Microsoft filed an amicus brief warning that the action "undermines military adherence to the rule of law." Five retired admirals and military leaders, including former CIA Director Michael Hayden, submitted supporting statements. Employees from Google, OpenAI, and Amazon have publicly affirmed support for Anthropic's position.
That coalition is remarkable. When your direct competitors' employees are backing your legal fight, the government's position has a credibility problem that extends beyond any single company.
The March 24 Hearing
A preliminary injunction hearing is scheduled for March 24, 2026. The Department of Justice has refused to commit to avoiding escalatory actions before then, meaning additional restrictions on Anthropic could arrive at any time.
The precedent here matters for every AI company that maintains usage policies. Content restrictions, safety guidelines, deployment limits - all of these become liabilities if the government can punish companies for keeping them in place. The incentive math is straightforward: comply with unrestricted government use, or lose access to federal contracts worth billions.
The commercial backdrop adds weight to the case. Claude's share of generative AI traffic has grown from 1.7% to 3.3% in recent months, though ChatGPT still holds 62% of the market. Anthropic is gaining ground commercially while simultaneously fighting the government over principles it considers non-negotiable. How the court rules on March 24 will shape whether AI companies can maintain safety boundaries when their largest potential customer demands otherwise.