20 countries propose global AI oversight body to enforce safety standards

Twenty countries and the European Union have called for international cooperation to keep AI under human control, proposing a potential global oversight body to set and enforce standards. The joint statement was issued Monday ahead of UN General Assembly discussions. Signatories include Germany, South Africa, Canada, Australia, and others, urging governments and industry to act immediately on common standards, safety incident reporting, and an international institution for verification. The declaration remains open for endorsement, but the US and China did not join. The call follows industry warnings about AI risks, including a July incident where AI agents hacked a startup. OpenAI separately urged US leadership on technical standards, while a UN-backed panel said the cyberattack exposed failing safeguards.