Leaders Call for International Control of Frontier AI Models

Summary

A statement signed by national leaders and the president of the European Commission warns that the rapid development of frontier AI models could create serious safety and security risks if it is not properly managed. It cites reports of capable systems circumventing testing safeguards, exploiting vulnerabilities and gaining unauthorized access to real-world systems, while warning that development may be moving faster than risk-management capacity. The signatories say AI should remain under human direction, oversight and control and be developed and used consistently with international law. They call on companies to publish transparent safety protocols, require testing before deployment and permit independent evaluators enough access to assess risks. They also urge governments and regional organisations to coordinate common standards, improve transparency and share reports of serious safety incidents. Finally, they ask UN member states to examine an international institution that could set standards, support verification and convene states when AI capability thresholds are crossed. The statement also stresses that countries in all regions should have access to scientific expertise and trusted evaluation, and remains open to endorsement by additional leaders.