The US government ordered Anthropic to withdraw its most capable AI models from public access. This directive targets high-performance systems to mitigate national security risks. Critics argue the move stifles open innovation and limits developer access. The decision forces a shift in how frontier models are deployed and monitored by federal regulators.