OpenAI backs four California bills, urges Congress to act on AI safety

OpenAI frames its policy push around what it calls a closing window — a limited period, which Greg Brockman has described as a "defenders window", when frontier AI can help defenders strengthen critical systems before powerful offensive capabilities become widespread. The company says governments face an analogous moment to establish durable safeguards before AI capabilities outpace the institutions responsible for governing them.

The announcement builds on a series of earlier positions: OpenAI has supported California's SB 53, New York's RAISE Act, Illinois's SB 315, and independent audits in frontier safety legislation under consideration in Massachusetts. It calls this state-by-state convergence "reverse federalism" — state safeguards creating a de facto national baseline that Congress can ultimately codify. The four new California bills address what OpenAI describes as complementary parts of the AI-safety ecosystem, from independent assessors to age assurance for companion chatbots to screening standards for gene-synthesis providers.

OpenAI also ties the push to its own internal work, including universal monitoring of full trajectories and a mandatory alignment-evaluation gate before broader internal deployment for its model Astra, and to Chief Scientist Jakub Pachocki's warning that the rapid rise of machine intelligence calls for "extreme caution." The company says it will slow or stop development when proceeding would pose an unacceptable safety risk. The outcome hinges on whether Congress moves before it adjourns, and on whether Governor Newsom signs the four California bills — a test of whether state action can build momentum for a federal framework.