Secretary-General António Guterres opened the first U.N. Global Dialogue on Artificial Intelligence Governance in Geneva on Monday, July 6, telling delegates that the technology is “advancing at runaway speed” and “faster than anyone – including the people building it – can keep up.” The forum is the first convened by the U.N. General Assembly in which all 193 member states hold a guaranteed seat on A.I., a structural departure from the invitation-only host-government summits at Bletchley Park, Seoul, Paris and New Delhi that defined the prior governance cycle.

The timing carried its own indictment. The day the Dialogue opened was also the day the deadline from the U.N.’s 2023 New Agenda for Peace lapsed, the deadline Guterres himself had set for a legally binding treaty on lethal autonomous weapons. No formal negotiations were ever begun. He used the podium to escalate anyway, calling such weapons “morally repugnant,” “politically unacceptable,” and declaring they “must be banned by international law.”

Guterres set out four priorities and rolled out a Child Safety Pledge for member states, framed around a comparison the U.N. hasn’t tried before on A.I.: “We do not let medicine reach a child until it is proven safe. We test every toy. Yet A.I. has reached our children – their learning, their friendships, their most private questions – before anyone asked what it would do to them.” Systems that detect distress in minors, he said, must “stop and connect them to real human support.”

General Assembly President Annalena Baerbock called the harms “sinister,” citing U.N. figures that 99 percent of deepfakes are sexual in nature and 96 percent target women and girls. Against the Secretary-General’s warning that “a patchwork of incompatible rules raises costs, divides the world – and protects no one,” the arithmetic of the room was harder to ignore. In September 2025, the White House Office of Science and Technology Policy declared that the United States rejects “centralized control and global governance of A.I.” On June 2, 2026, a U.S. executive order established a voluntary framework for frontier-model safety reviews, the domestic answer to the multilateral question Geneva was convened to ask.

The I.T.U.’s Doreen Bogdan-Martin, folding the summit into the concurrent WSIS Forum and AI for Good programming, branded the convergence “Geneva Digital Week.” The second Dialogue is scheduled for New York in May 2027. That’s the operative fact. The U.N. has secured a permanent seat for every state at the table it built, and simultaneously watched its most concrete deliverable, the autonomous-weapons treaty, expire on the table’s opening day.

Sources