A Defense Department memo dated March 6, 2026, reportedly directed every military component to purge Anthropic AI products from its systems and networks. The Pentagon classified Anthropic as a “supply chain risk”, a designation that, under procurement law, triggers mandatory removal.
According to reporting on the directive, the 180-day removal window extends to defense contractors, who must certify compliance with the ban. That puts the approximate removal deadline at September 2, 2026, though that date is approximate, pending review of whether the memo specifies calendar or business days.
The reported basis for the ban is a policy dispute, not a security failure. According to reporting on the dispute, Anthropic refused to grant blanket authorization for its tools to be used in autonomous weapons systems or for mass surveillance. The government reportedly offered assurances it wouldn’t use the technology for those purposes, assurances Anthropic considered insufficient without contractual guardrails. The memo followed.
Anthropic is challenging the supply chain risk designation in court, arguing the government retaliated against its protected viewpoint on AI safety, a First Amendment framing that, if sustained, would set significant precedent for how the government can use procurement exclusions against companies over policy disagreements.
All claims in this brief are sourced from reporting provided via source hints; no working source URLs were confirmed during the Filter’s verification stage. This brief will be updated once Wire amendment confirms working URLs.