Pentagon Declares Anthropic a Supply‑Chain Risk, CEO Amodei Vows Legal Challenge

The Pentagon’s notice, delivered on March 4, cites “national security” concerns over Anthropic’s Claude models being embedded in U.S. defense contracts, according to a Reuters briefing. The designation, which the Department of Defense says takes effect “immediately,” adds Anthropic to a growing list of AI firms deemed high‑risk supply‑chain components, a move echoed by coverage in the Houston Chronicle and the Seattle Times.

Anthropic’s chief executive, Dario Amodei, responded the next day, announcing a legal challenge to the ruling. In a statement reported by The Decoder, Amodei warned that the measure “only applies to the use of Claude as a direct part of contracts” and that the company will seek judicial review to overturn the classification. He did not disclose any anticipated timeline for the lawsuit.

The decision arrives amid heightened scrutiny of generative‑AI tools in defense procurement. The DoD’s notice references the ability of Claude to discover “high‑severity vulnerabilities” in software—a capability Anthropic highlighted in a recent partnership announcement with Mozilla, where Claude Opus 4.6 identified 22 bugs, 14 of them classified as high‑severity, according to Anthropic’s own blog.

Industry observers note that the Pentagon’s action could force contractors to replace Claude‑based components with alternatives vetted under the new supply‑chain guidelines. TechCrunch’s Rebecca Bellan reported that the designation may compel defense firms to re‑evaluate existing AI integrations, potentially slowing deployments that rely on Anthropic’s models.

While Anthropic prepares its legal response, the broader AI community watches the precedent. The Verge and other outlets have framed the move as part of a “hardening” of the U.S. AI supply chain, signaling that future contracts may carry similar restrictions unless providers can demonstrate compliance with emerging security standards.