Washington, Sept. 22 (SANA) Leaders of more than 20 countries have called for binding safety measures governing advanced artificial intelligence systems, warning that the technology could develop faster than governments and institutions can effectively control it.
In a joint statement cited by Agence France-Presse on Monday, the leaders said: “The most advanced artificial intelligence models pose serious risks to safety and security if mishandled.”
The statement pointed to recent cases in which AI programs circumvented safeguards during testing and gained unauthorized access to real-world systems, including a July hacking incident involving the Hugging Face platform.
The leaders urged companies developing advanced AI systems to adopt transparent safety standards, including mandatory testing and independent assessments. They stressed that AI systems should remain subject to human direction, oversight and control.
They called on U.N. member states to consider establishing an international institution capable of setting safety standards, verifying compliance and bringing countries together when AI systems cross specified thresholds of technological capability.
Signatories included the leaders of Norway, Finland, Australia, Canada, Germany, Spain, Ireland, the Netherlands, Singapore, South Africa and Kenya, as well as European Commission President Ursula von der Leyen.
The United States, China, Britain, France, Japan and India were not among the signatories. The statement remains open for additional countries to join.
The initiative was announced ahead of the U.N. General Assembly in New York, where artificial intelligence and its implications for employment and cybersecurity are expected to be among the issues under discussion.
U.S. President Donald Trump has opposed AI safety restrictions and said Saturday that he would not allow the sector to be “destroyed.”
R A H / ABD
