OpenAI's proposal for binding global AI standards

OpenAI published a proposal calling on the US government to develop binding international safety and security standards for frontier AI systems, reversing the company's earlier opposition to government-imposed AI regulation. The proposal covers capability evaluations, risk assessment, human oversight, incident reporting, and recursive self-improvement. OpenAI calls for shared technical standards on RSI across the industry and states that fully autonomous RSI should not be pursued until safety can be demonstrated. It recommends the US Center for AI Standards and Innovation collaborate with AI safety institutes in the UK, France, and Singapore, and calls for secure channels among countries to share emerging AI threats. The proposal would track gains in RSI-relevant capabilities and the degree to which lab research is performed autonomously, with thresholds triggering immediate human review, and warns that highly autonomous AI research could become too opaque for effective human supervision. Sam Altman stated that 'people outside the AI labs should have a real say in how this technology develops, and a clear way to judge if it's happening safely,' and that standards should 'help prevent the concentration of power, including by making sure new companies and open-model companies can compete.' OpenAI Chief Scientist Jakub Pachocki separately urged governments to make international AI coordination a top priority. The proposal was released as separate US-China talks were underway on a notification mechanism for AI incidents of national-security significance. Anthropic CEO Dario Amodei separately put forward a plan to moderate the pace of frontier AI development and sought a limited antitrust exemption allowing tech companies to coordinate on safety.