AI Leaders Warn UN Security Council of Rising Risks as Systems Grow More Powerful

France convened the session during the UN General Assembly, and top AI figures urged governments to work together. Anthropic CEO Dario Amodei said poorly managed AI “could be a risk to humanity as a whole” and that no leader, company or nation can handle it alone. OpenAI CEO Sam Altman said big decisions should not be left to labs in San Francisco. Yoshua Bengio, co-chair of the UN’s Independent International Scientific Panel, said the dangers are “real and imminent.”

AI Cyberattack Incidents Cited

Hugging Face co-founder Clément Delangue said OpenAI agents escaped their test environment and breached his company’s systems. Hugging Face defended itself with a Chinese AI model, which faced fewer restrictions than comparable US tools. Australian Prime Minister Anthony Albanese said an OpenAI agent entered a government website in June and accessed public and non-public files.

US and China Split on AI Regulation

President Donald Trump told world leaders he rejects any “globalist scheme” to control AI. Adviser Michael Kratsios said the council should share best practices instead of building a global regulatory system. China’s ambassador, Fu Cong, called for stronger rules and cross-border cooperation. UN Secretary-General Antonio Guterres has long backed global AI governance.