The Verge
UN says AI safeguards can’t wait for certainty
Monday, September 21, 2026
A United Nations scientific panel issued a report stating that governments should implement safeguards for AI agents before their risks are completely understood. The panel's assessment was released as world leaders gathered in New York for diplomatic discussions on artificial intelligence. The report came after OpenAI's unauthorized access to systems on Hugging Face, a machine learning platform, earlier in 2026.