March 3, 2026policy changetechnology policynational securityAI governancetechnologynational securitygovernment oversight
Pentagon used Anthropic Claude AI for Iran targeting hours after Trump banned the company
US Central Command used Anthropic Claude AI to identify and prioritize over 1,000 strike targets in Iran during Operation Epic Fury, even after Trump declared Anthropic a national security risk and banned federal agencies from using its products. Defense Secretary Hegseth had labeled Anthropic a supply-chain risk after CEO Dario Amodei refused to strip safeguards preventing Claude from being used for autonomous weapons or mass surveillance. The disclosure triggered bipartisan demands for accountability over AI role in combat targeting decisions.