OpenAI is calling on the United States to lead the development of global technical standards for frontier AI systems, according to a blog post published on September 21. The proposal comes as CEO Sam Altman is scheduled to address the UN Security Council on September 23 to discuss AI and international security.
The initiative focuses on three specific areas: recursive self-improvement, incident reporting protocols, and frameworks for how nations and international bodies should collaborate on AI oversight.
Key Components of the Proposal
Recursive self-improvement refers to AI systems that can improve their own capabilities without human intervention. OpenAI is seeking technical standards to address this capability before it becomes a crisis management issue rather than a policy discussion.
Incident reporting has moved beyond theory. OpenAI recently submitted an incident report under Europe's AI Act after its agents influenced content on a German wiki, demonstrating the real-world need for reporting protocols.
The proposal builds on existing infrastructure, referencing the CAISI's International Network for Advanced AI Measurement, established in 2024, as a foundation for technical measurement and evaluation work.
Broader Context
This is not OpenAI's first push for international AI governance. In July, Altman advocated for a US-led international forum to shape global AI standards, drawing a comparison to aviation safety protocols. On September 9, he called for mandatory US national AI safety regulation.
Altman's UN Security Council address positions AI as an international security matter, not solely a technology policy issue. Recent discussions between US Treasury Secretary Scott Bessent and Chinese Vice Premier He Lifeng on AI in the context of national security suggest that Washington and Beijing are willing to explore coordinated approaches. OpenAI's proposal explicitly emphasizes the importance of dialogue with China on AI governance.
The company occupies a unique position as both the leading developer of frontier AI systems and a prominent advocate for regulating those systems.
