UN Chief Warns AI Outpacing Governance, Calls for Global Dialogue

UN Secretary-General António Guterres opened the first Global Dialogue on AI Governance in Geneva on Monday telling 193 nations that AI is already outpacing the institutions meant to govern it—and that humanity is running an experiment on itself "without a plan, and without consent."

"Artificial intelligence is advancing at runaway speed,” Guterres said in the opening of his keynote. “A technology that can reshape economies, transform the world of work, sway elections, and tilt the balance of security is being deployed faster than anyone—including the people building it—can keep up."

The room included all 193 UN member states, convened in Geneva at the first Global Dialogue on AI Governance, a multilateral attempt to govern a technology that has already outrun everyone trying to do it.

The meme used by Guterres to explain the risks of ungoverned AI was vibe coding—a term coined by Andrej Karpathy, founding member of OpenAI and former director of AI at Tesla, to describe programming by feel: tell AI what you want, let it handle the rest, don't look too closely. Merriam-Webster recently added it to its dictionary.

Guterres acknowledged vibe coding "can do wonders." as more people trust AI-built products.

"But we cannot vibe-code the truth,” he said. “We cannot vibe-code the future of humanity."

The numbers behind that line weren't soft. Guterres put the internet at 15 years to reach a billion people and AI at two. He described current systems as "no longer tools awaiting instruction—they are writing code, acting online, and making choices with less and less human oversight."

"Our institutions were built to govern machines that follow commands. They are not ready for machines that decide."

The dialogue was born from the 2024 Global Digital Compact, which established international AI governance as a UN mandate for the first time. Its opening session two years later also received the preliminary report of the Independent International Scientific Panel on Artificial Intelligence—40 scientists from 140 countries—which published last week the finding that nobody can currently guarantee AI won't cause catastrophic harm.

Speed was one of three warnings Guterres drew from the panel. The second was power: computing, data, and talent concentrated in a handful of companies and countries, most of the world locked out of decisions already shaping it, something that also worries other AI experts like Yann Lecun, Andy Kowinski and Yoshua Bengio.

The third was truth—a machine-enabled lie now persuades as effectively as a verified fact, steadily eroding what Guterres called "the integrity of our information ecosystem."

Child guinea pigs and killer robots

Among Guterres’ concrete proposals are an AI Child Safety Pledge requiring companies to prove safety through independent testing before any AI reaches children, maintain zero tolerance for the generation of child sexual abuse imagery, and connect distressed children to real human support rather than leaving them alone with a chatbot. "No child should be a guinea pig for unregulated AI."

And then there were the killer robots. Guterres called lethal autonomous weapons—machines that select and kill a target without human judgment—"morally repugnant" and demanded a ban by international law.

States are already at the discussion table. He did not suggest they take their time.

The Dialogue reconvenes in New York in 2027. Guterres also called on the General Assembly to create a Global Fund for AI focused on computing access for developing countries, and challenged every major AI company to run all data centers on renewable energy by 2030—the year he estimates those facilities will outpace all but five nations in electricity consumption.