AI Safety

Dispatch from Anthropic v. Department of War Preliminary Injunction Motion Hearing

Federal judge questions government's retaliation against Claude creator for refusing autonomous weapons clause.

Deep Dive

In a landmark hearing at San Francisco's Phillip Burton Federal Building, Anthropic PBC faced off against the U.S. Department of War before Judge Rita F. Lin. The core conflict emerged when the Department attempted to renegotiate its contract with Anthropic to allow "all lawful uses" of the Claude language model. Anthropic refused to compromise on two ethical red lines: prohibiting autonomous weapons systems and mass surveillance of American citizens. Following this refusal, the government took three contested actions: banning other federal agencies from using Claude, announcing a secondary boycott prohibiting federal contractors from doing business with Anthropic, and formally designating the company as a supply chain risk.

During the preliminary injunction hearing, Judge Lin directly challenged the government's legal position. She focused on a February 27th social media post by Secretary Pete Hegseth declaring a broad commercial boycott, asking defense counsel if the Secretary had the authority for such a directive. Deputy Assistant Attorney General Eric Hamilton argued the post merely announced an ongoing process and wasn't legally binding, a claim Judge Lin met with skepticism, stating, "You're standing here saying, we said it, but we didn't really mean it." The judge also questioned the practical scope, asking if a military toilet paper supplier using Claude for coding would be affected. Hamilton conceded it would not, contradicting the Secretary's public statement. The hearing's outcome will determine if Anthropic gets temporary relief from the government's actions while the full case proceeds.

Key Points
  • Contract dispute centers on Anthropic's refusal to allow Claude AI for autonomous weapons or mass surveillance, which were non-negotiable 'red lines'.
  • Government retaliated with three actions: a cross-agency ban on Claude, a contractor boycott, and a supply-chain risk designation.
  • Judge Lin expressed clear skepticism of the government's defense, particularly regarding the authority and intent behind a public boycott declaration.

Why It Matters

Sets a critical precedent for AI companies enforcing ethical use policies against government pressure and defines limits of state retaliation.