Anthropic sues Pentagon to fight unprecedented "supply chain risk" label
What's the deal? AnthropicDealroom has a profile for this one. Try Dealroom → filed two federal lawsuits on March 9, 2026, challenging the US Department of DefenseDealroom has a profile for this one. Try Dealroom →'s decision to label it a "supply chain risk" — a designation that effectively bars military contractors from using its technology.
The suits, filed in California and the DC Circuit Court of Appeals, argue the Pentagon overstepped its legal authority and violated Anthropic's First Amendment rights by retaliating against the company for its stated views on AI safety.
The dispute stems from a collapsed $200 million contract negotiationDealroom has a profile for this one. Try Dealroom →. Anthropic refused to agree that the Pentagon could use its Claude AI for any lawful purpose, insisting on guardrails against mass domestic surveillance and fully autonomous weapons.
Defense Secretary Pete HegsethDealroom has a profile for this one. Try Dealroom → responded by formally designating Anthropic a supply chain risk — a label previously reserved for firms with ties to China or Russia, and never before applied to an American company.
Why now? Talks broke down for good in early March 2026, partly after an internal memo by Anthropic chief executive Dario AmodeiDealroom has a profile for this one. Try Dealroom → — accusing the administration of opposing the company because it had not given "dictator-style praise to Trump" — was leaked to the press. The Pentagon formally notified Anthropic of the designation shortly after, leaving the company with no path forward other than litigation.
What could go wrong? Legal experts say Anthropic faces an uphill battle. Government contracting law gives the Pentagon broad discretion over supply chain risk designations, with limited grounds for appeal.
Anthropic's stronger argument may be that it was singled out arbitrarily — OpenAI struck a deal with the Pentagon around the same time, reportedly securing similar safety protections while agreeing to the "all lawful uses" standard.
Meanwhile, the designation is already having commercial consequences. Several Anthropic customers are exploring alternatives, and rivals including OpenAI and xAIDealroom has a profile for this one. Try Dealroom → have moved quickly to fill the gap on classified Pentagon systems.
The signal: This case will define how much control AI developers can retain over how their models are used once they enter government contracts — a question with no precedent. The Pentagon's position is that a private company cannot set policy for the US government; Anthropic's is that safety constraints are a legitimate condition of doing business.
Whichever side prevails, the outcome will shape the terms on which every AI lab engages with the defence sector for years to come.
Sources:
Anthropic
The Washington Post
Reuters
Politico
Wired
Forbes
The Guardian
Bloomberg
The New York Times
TechCrunch
BBC
CNN
J.V.
