Anthropic refused the Pentagon’s demand to remove AI safety restrictions from its $200 million military contract. The consequences arrived within hours.
President Trump directed federal agencies to cease using Anthropic’s products on Friday afternoon, with the Pentagon given six months to phase out Claude from classified networks. Defense Secretary Pete Hegseth then designated Anthropic a supply chain risk, a classification previously reserved for companies suspected of ties to foreign adversaries. It is the first time the United States government has applied this designation to an American company.
The standoff centered on two specific guardrails Anthropic CEO Dario Amodei refused to remove: prohibitions on mass domestic surveillance of Americans and fully autonomous weapons systems. The Pentagon wanted Claude available for “all lawful purposes” without restriction. Amodei responded that “no amount of intimidation or punishment from the Department of War will change our position” and announced Anthropic would challenge the designation in court.
OpenAI CEO Sam Altman announced a classified deployment deal with the Pentagon late Friday, claiming his agreement includes the same two restrictions Anthropic had fought to maintain, plus a third ban on high-stakes automated decisions like social credit systems. OpenAI stated its deal “has more guardrails than any previous agreement for classified AI deployments, including Anthropic’s.”
The practical fallout extends beyond the Pentagon. Every defense contractor doing business with the military is now prohibited from conducting commercial activity with Anthropic. Legal and policy analysts noted the designation could take years to resolve in court, with the business damage potentially irreversible regardless of the legal outcome.
Source: Anthropic Official Statement | CNN | Feb 27, 2026