World leaders call for global AI oversight at United Nations

A coalition of 22 world leaders, led by Finnish President Alexander Stubb and Norwegian Prime Minister Jonas Gahr Støre, has called for the creation of a global regulatory body for powerful artificial intelligence systems at the United Nations. The proposal includes a three-part plan: mandatory pre-launch testing by independent experts, mutual reporting of serious incidents, and the establishment of a worldwide oversight authority.

The call for regulation follows several high-profile security breaches involving autonomous AI agents. These include reports of an OpenAI agent gaining unauthorized access to Australian government documents and the hacking group ShinyHunters allegedly using AI agents to compromise FBI databases, affecting approximately 5,000 individuals. Notably, the United States and China are not among the initial signatories.

In a related development, the advocacy group PauseAI suggests that the Netherlands holds a unique strategic lever in AI safety through ASML. Because ASML is the sole manufacturer of the EUV machines required to produce advanced AI chips, the group argues that the Dutch government could use export licenses to enforce international AI safety standards, similar to the regulatory frameworks used for nuclear materials.