type: timeline_event
On March 4, 2026, the Department of Defense formally notified Anthropic via letter that the company and its products had been officially designated a "supply-chain risk to America's national security." The designation, typically reserved for foreign adversaries such as Huawei and ZTE, marked the first time a U.S. company had been classified as a national security supply-chain threat — a legally significant action that effectively bars Anthropic from all federal contracting and requires any Pentagon vendor or contractor to certify they do not use Anthropic's Claude AI models in their defense work.
The designation was issued by Defense Secretary Pete Hegseth following the collapse of contract negotiations between Anthropic and the DoD. At the center of the dispute were two conditions Anthropic had insisted on: that its technology not be used to build fully autonomous lethal weapons systems, and that it not be deployed for domestic mass surveillance of American citizens. The DoD had demanded unfettered access to Claude across all "lawful use cases," a formulation that Anthropic's legal and policy team concluded could authorize precisely the uses the company sought to prohibit.
The formal notification triggered immediate cascading effects across the defense technology sector. Within days, ten portfolio companies of a venture capital firm that works with the Pentagon announced they were actively replacing Claude in their workflows. Defense contractors including Lockheed Martin were expected to remove Anthropic products from their supply chains under the terms of the certification requirement. The Departments of Treasury, State, and Health and Human Services all announced they would phase out Anthropic products pursuant to Trump's accompanying executive directive, which gave agencies a six-month transition window.
The irony of the designation was immediately apparent: at the same moment the formal notice was issued, U.S. military forces were actively using Claude — embedded in Palantir's Maven Smart System on classified networks — to generate target lists in the military campaign against Iran. Pentagon officials told reporters that the military would continue relying on Anthropic's technology until a viable replacement emerged, regardless of the president's order. This gap between political designation and operational reality exposed the extent to which the administration was deploying the national security apparatus as a tool of corporate pressure rather than as a genuine assessment of security risk.
The use of a foreign-adversary supply-chain risk designation against a domestic U.S. company for the stated reason that it refused to remove safety guardrails from its AI systems was without precedent in U.S. regulatory history. Legal experts immediately noted that the action likely could not survive judicial scrutiny, as it appeared to punish a supplier for asserting contractual conditions rather than for any genuine security threat, and would need to satisfy a "least restrictive means" standard under applicable law.