AI safety

OpenAI has published a blueprint for U.S. frontier AI governance, focusing on national rules, CAISI, and broader resilience planning.