OpenAI outlines blueprint for federal AI safety governance

OpenAI has released a blueprint for a federal framework to govern increasingly capable AI systems in the United States. The proposal focuses on building durable institutions and coordinating safety policy as frontier AI advances. The blueprint highlights three priorities: creating a national framework that builds on emerging statelevel safety laws, strengthening CAISI as the federal government’s main frontier AI safety body, and expanding a broader resilience plan across government to address national security and public safety risks. The company points to recent actions by states such as California, New York, and Illinois, along with a new White House executive order on advanced AI, as signs that the policy environment is moving toward more coordinated oversight. OpenAI says the federal government should now develop a framework that can adapt as the technology changes.