Anthropic Sues US Government Over 'Supply Chain Risk' Designation

The legal confrontation between Anthropic and the US government represents a watershed moment in the relationship between Silicon Valley’s artificial intelligence sector and the Department of Defense. By filing suit in a California federal court, Anthropic is challenging the Trump administration's unprecedented decision to label a domestic technology leader as a 'supply chain risk'—a designation historically reserved for foreign adversaries like Huawei or ZTE. This move by the Pentagon, led by Defense Secretary Pete Hegseth, appears to be a direct retaliation against Anthropic’s refusal to grant the military unfettered access to its Claude AI models without ethical guardrails.

At the heart of the dispute is a fundamental disagreement over the 'terms of service' governing national security contracts. Anthropic alleges that Secretary Hegseth demanded the removal of long-standing clauses that prohibit the use of its technology for lethal autonomous warfare and the mass surveillance of American citizens. Anthropic’s refusal to comply led to its blacklisting, which the company describes as an 'unlawful' use of executive power intended to punish protected speech and corporate policy. The administration’s rhetoric, delivered through White House spokeswoman Liz Huston, frames the conflict as a battle against 'woke' corporate overreach, signaling a shift toward a more coercive 'America First' approach to defense procurement.

“

This move by the Pentagon, led by Defense Secretary Pete Hegseth, appears to be a direct retaliation against Anthropic’s refusal to grant the military unfettered access to its Claude AI models without ethical guardrails.

The implications for the broader defense-tech ecosystem are profound. If the government’s 'supply chain risk' designation is upheld, it provides the executive branch with a powerful tool to force compliance from private technology providers. This could create a chilling effect across the AI industry, forcing companies like OpenAI or Meta to choose between maintaining their safety frameworks and maintaining their eligibility for lucrative government contracts. Conversely, an Anthropic victory would reinforce the right of private entities to set ethical boundaries on the deployment of their intellectual property, even within the context of national defense.