Anthropic sues US government over ‘supply chain risk’ label

TL;DR: Anthropic has filed a lawsuit against the US government after the Pentagon labelled it a “supply chain risk” for refusing to remove restrictions on lethal autonomous warfare and mass surveillance from its military contracts. Google and OpenAI employees have filed a court brief supporting Anthropic’s position.

Anthropic, the AI company behind Claude, has taken the Trump administration to court in what it calls an “unprecedented and unlawful” retaliation for maintaining safety restrictions on its military AI contracts.

The dispute

The row centres on Anthropic’s refusal to give the US military unrestricted use of Claude. Defence Secretary Pete Hegseth demanded the company remove all usage restrictions, despite contract limitations on lethal autonomous warfare and mass surveillance of Americans having been standard terms since Anthropic first worked with government agencies in 2024.

Anthropic says it was close to negotiating revised contract language when President Trump publicly called the company’s leadership “left wing nut jobs” and directed all government agencies to stop using its tools. Hegseth followed by designating Anthropic a “supply chain risk”, effectively banning Claude from government use and prohibiting any government contractor from using Anthropic products.

Industry response

The fallout has been significant. Microsoft, Google, and Amazon confirmed last week they would continue using Claude outside defence work. Nearly 40 employees from Google and OpenAI filed a court brief supporting Anthropic, stating they were “united in the conviction” that frontier AI systems need guardrails around mass surveillance and autonomous weapons.

OpenAI CEO Sam Altman admitted to rushing through a new Defence Department contract in the wake of Anthropic’s fallout with the government.

Anthropic claims “hundreds of millions of dollars” in contracts are now at risk and is seeking a court declaration that Trump’s directive exceeds presidential authority.

Looking forward

Legal experts expect a protracted battle. Carl Tobias of the University of Richmond School of Law predicts the Trump administration will take “a scorched earth” approach, with the case likely reaching the Supreme Court. The outcome could set a precedent for whether AI companies can maintain safety restrictions when selling to governments.