The most consequential development in artificial intelligence governance this year was not a joint declaration, a regulatory breakthrough or a geopolitical confrontation. It was a quiet, procedural vote: the United Nations General Assembly’s recent endorsement of a new scientific panel tasked with assessing AI’s risks and opportunities. That it passed with little fanfare, during the same week that attention was focused on world leaders gathered in New Delhi for the fourth international AI summit since ChatGPT’s debut in 2022, says something about where things stand.
The U.N. body, formally known as the International Scientific Panel on Artificial Intelligence, has a telling backstory. The United States voted against its establishment, even as 117 other countries voted in favor, among them U.S. allies across Europe and Asia, alongside Russia, China and much of the developing world. That alignment is itself a map of the current moment. Scientific panels derive their influence not from formal authority but from something harder to contest: the slow accumulation of global political consensus.
Climate diplomacy offers the clearest precedent. When the Intergovernmental Panel on Climate Change was created in 1988, it did not immediately set out to regulate carbon emissions or negotiate global agreements. Instead, it framed the problem. Over time, the IPCC’s assessments hardened into the intellectual architecture that diplomats and markets alike learned to navigate.