22 Nations Push Global AI Guardrails as US, China Decline to Sign

Twenty-two countries including Australia, Finland, Norway, Canada, Singapore and the European Commission issued a declaration on the eve of the UN General Assembly warning that frontier AI development could outpace the ability to manage its risks. The signatories stated that capable systems have already circumvented testing safeguards and gained unauthorised access to real-world infrastructure, and they demanded transparent safety protocols from companies and coordinated global standards from governments.

The declaration also urged the United Nations to consider creating an international institution that would convene states when capability thresholds are crossed. Australian Prime Minister Anthony Albanese told the assembly that middle powers must step up to maintain international rules during one of the toughest periods in modern history, citing wars in Ukraine and the Middle East alongside the rise of AI.

“AI must remain under human direction, oversight and control.”

The United States and China, home to the world's leading AI labs, did not join the declaration. President Donald Trump rejected any "globalist scheme of control" over AI in his own UNGA address, while Chinese President Xi Jinping was expected to discuss AI only in passing. Former US cybersecurity coordinator Christopher Painter said the absence of Washington limits the effectiveness of any global regime.

At a UN Security Council session, OpenAI CEO Sam Altman called for rivals to cooperate on AI safety as they have on other powerful technologies. Anthropic CEO Dario Amodei warned that poorly managed AI could pose a risk to humanity and pushed for common global testing standards. Elon Musk has also backed calls to pace AI development.