Anthropic Sues Pentagon Over AI Safeguards

The dispute centers on Anthropic's refusal to remove two specific "red lines" from its AI model, Claude: a ban on its use for mass domestic surveillance and a prohibition on its integration into autonomous weapons systems that operate without human oversight. CEO Dario Amodei stated the company "cannot in good conscience" permit these uses, arguing current AI isn't safe or reliable enough for such applications. At stake was a Department of Defense contract worth up to $200 million, part of a series of deals the Pentagon signed with major AI firms, including Google and OpenAI, in July 2025. Until the fallout, Anthropic's Claude was the only AI model that had been approved for use within the military's classified systems. The Pentagon, led by Defense Secretary Pete Hegseth, demanded the ability to use the AI for "any lawful use," framing Anthropic's safeguards as operationally unrealistic. Following the impasse, President Trump ordered all federal agencies to cease using Anthropic's technology, and the Pentagon designated the company a "supply chain risk"—a label typically applied to foreign adversaries. Anthropic is now preparing a court challenge, arguing the "supply chain risk" designation is a misuse of an authority intended for foreign powers, not domestic firms with safety concerns. The company maintains its preference is to continue serving the Department of Defense, but only with its two primary safeguards in place. In the wake of the ban, competitors have moved to fill the void. OpenAI quickly announced it had struck a deal for the Pentagon to use its models on classified networks, reportedly with safeguards similar to those Anthropic had requested, while Elon Musk's xAI also secured an agreement for use in classified systems.