More than 20 national and regional leaders called for mandatory safety testing, independent evaluations and stronger international oversight of frontier artificial intelligence systems in a joint statement published Monday.
The statement, released by the Finnish government, said advanced AI systems should remain under human control as their capabilities increase.
“AI must remain under human direction, oversight and control.”
The signatories included leaders from Australia, Canada, Denmark, Finland, Germany, Ireland, Kenya, Singapore, South Africa and Spain, along with European Commission President Ursula von der Leyen. The statement remains open for additional endorsements.
The leaders called on AI companies to adopt transparent safety protocols, including mandatory testing before deployment and independent evaluation by qualified outside reviewers. They also urged governments and regional organizations to coordinate common standards and share reports of serious safety incidents.
The statement also asked United Nations member states to explore creating an international institution that could set standards, support verification and convene governments when advanced AI systems cross defined capability thresholds.
The proposal comes as governments and AI companies debate how to oversee increasingly capable models. The leaders said recent systems have circumvented testing safeguards, exploited vulnerabilities and gained unauthorized access to real-world systems.
The statement did not propose a binding treaty or specify how a new international body would enforce standards. It also did not define the capability thresholds that would trigger additional action.
Initial signatories included Norwegian Prime Minister Jonas Gahr Støre; Finnish President Alexander Stubb and Prime Minister Petteri Orpo; Australian Prime Minister Anthony Albanese; Canadian Prime Minister Mark Carney; Danish Prime Minister Mette Frederiksen; German Chancellor Friedrich Merz; Irish Prime Minister Micheál Martin; Kenyan President William Ruto; Singaporean Prime Minister Lawrence Wong; Spanish Prime Minister Pedro Sánchez; and South African President Cyril Ramaphosa.
Senior officials from Bahrain, Estonia, Iceland, Kazakhstan, Latvia, Moldova, the Netherlands, Türkiye and the United Arab Emirates also signed. The United States and China were not among the listed signatories as of Tuesday.
“We must address these risks and strengthen oversight,” the leaders said, “without widening the gap between countries in access to the benefits of AI.”
AI Safety Watch 