OpenAI urges US-led global standards for frontier AI

OpenAI's proposal follows its earlier statement of three goals for navigating advanced AI: building an automated AI researcher, delivering scientific and economic benefits, and empowering individuals with a personal AGI. The company says fully autonomous recursive self-improvement is not happening today and should not be pursued unless it can be done safely, citing its disclosed Hugging Face Incident as a preview of risks that could worsen without safeguards.

The standards push builds on existing institutions. OpenAI points to the Center for AI Standards and Innovation (CAISI), state laws, and a federal framework as helpful, and notes that the US helped create the International Network for Advanced AI Measurement, Evaluation, and Science in 2024 to cooperate on measurement and evaluation. It also lists standards bodies such as ISO, the Frontier Model Forum, Agentic AI Foundation, and Open Secure AI Alliance as partners.

The core tension is between national authority and cross-border coordination. OpenAI argues that a US-led effort can avoid a fragmented, uneven, and conflict-ridden system while still leaving governments free to adopt standards as they see fit. Whether other countries and labs buy into shared measurement and incident-reporting protocols—and whether upcoming US-China talks produce cooperation—will determine if the approach gains traction.