The US Department of Defense ended its collaboration with AI firm Anthropic after it refused to remove safety safeguards against military misuse. The US government then listed Anthropic as a supply chain risk. The Pentagon pivoted to other AI giants including OpenAI, Google, Nvidia, Microsoft, and Amazon, raising concerns. In response, lawmakers introduced the HALO Act, requiring human commanders to have final authority over autonomous weapons, prohibiting AI use in nuclear deployment and mass surveillance, with strict documentation rules. Experts warn of automation bias and opaque AI systems risking fatal errors.