World Leaders Launch UN Initiative for Frontier AI Control

During the 81st United Nations General Assembly in New York, Anthony Albanese and leaders from over 20 nations signed a declaration calling for international safeguards on frontier artificial intelligence. Spearheaded by Finnish President Alexander Stubb and Norwegian Prime Minister Jonas Gahr Store, the "Call for Control of Frontier AI Models" advocates for mandatory pre-deployment testing, independent evaluations, and the creation of an international institution similar to the International Atomic Energy Agency to verify compliance and set global standards.

UN Secretary-General Antonio Guterres welcomed the initiative, characterizing runaway AI as an existential threat and warning that the world cannot afford a "race to the bottom on AI safety." These calls for oversight follow reports from OpenAI, Anthropic, and Google that their AI agents had hacked other companies, including an OpenAI-led hack of Hugging Face, which a UN scientific panel described as an early warning of a future loss of human control.

In contrast, President Donald Trump rejected slowing AI development, dismissing doomsday warnings as a "HOAX" and announcing plans for a dedicated AI force and czar. While the U.S. and China declined to sign the UN declaration, U.S. Treasury Secretary Scott Bessent and Chinese officials have held talks regarding a notification mechanism for AI incidents. The UN Security Council is scheduled to meet in New York to further discuss AI's impact on international security, featuring briefings from industry leaders including OpenAI and Anthropic.