OpenAI asks the U.S. government to coordinate international rules for assessing the risks of artificial intelligence

OpenAI is calling on the United States government to work with national regulators to develop international standards for assessing and classifying the risks associated with artificial intelligence. The proposal was submitted on Monday, amid growing concerns about the technology’s rapid development and use.

The American start-up recommends using the International Network for Advanced AI Measurement, Evaluation and Science, created in 2024 by the Center for AI Standards and Innovation, an organization associated with the U.S. government. The network includes the United States, the European Union, France, Canada, the United Kingdom, Japan and South Korea.

OpenAI notes that the standards would not require approval before an AI model could be marketed. Each government would decide whether and how to incorporate them into national legislation. The company proposes that the criteria be developed in consultation with industry representatives, independent experts and academic researchers.

Calls for closer international coordination have intensified in recent months, but without concrete results. The Donald Trump administration is reluctant to accept new regulatory commitments and views AI development as a global competition that the United States must win through a permissive, deregulation-oriented approach.