OpenAI Urges US Leadership on Global Frontier AI Standards

OpenAI has urged the United States to spearhead an international effort to set technical standards for frontier artificial intelligence, including systems capable of recursively improving their own capabilities.

The company said shared safety and security standards are growing more urgent as frontier AI is developed and deployed across borders. Divergent national rules, it warned, could fragment evaluations and reporting, making it harder to tackle risks that do not stop at frontiers.

In its view, the US is well placed to lead, citing its position at the forefront of AI development and its influence across global networks in finance, trade, defence, technology and information systems. It called for Washington to work with other countries to build common technical standards for frontier AI, including for recursive self-improvement (RSI).

RSI refers to AI systems taking on more of the work of developing successive model generations. OpenAI said this could accelerate AI research beyond the collective ability to understand its progress, assess risks and keep meaningful human oversight. An uneven distribution of frontier AI development and expertise across countries could further complicate those efforts.

The company stressed that fully autonomous RSI is not happening today and should not be pursued unless it can be done safely, with human control and oversight preserved. Without adequate safeguards, it warned, RSI could leave humans unable to oversee research processes they no longer understand, making AI more dangerous and less aligned.

OpenAI also outlined a three-part strategy for advancing towards artificial general intelligence: building an automated AI researcher to help solve alignment challenges, accelerating scientific and economic gains from more capable systems, and ultimately giving individuals access to personal AGI. It said automated AI research could lower the cost of advanced intelligence and bring benefits to people worldwide.

On the stakes, the company said leading now will determine whether the US shapes the global AI framework or watches a fragmented, uneven and conflict-ridden system take hold — and whether advanced AI strengthens the networks the US and its allies already share.