UN chief calls for worldwide AI safety cooperation amid growing risks

When António Guterres stepped up to the podium in New York this week, his message was clear, if a little uneasy: the rapid sprint of artificial‑intelligence research is outpacing the world’s ability to understand, let alone control, the risks.

“We can’t afford a race to the bottom on AI safety,” the U.N. chief told reporters, his tone a mixture of urgency and the weary patience of someone who has heard the same warnings for months. He wasn’t just talking about any technology – he meant the next‑generation systems that can write code, generate deepfakes, or even make strategic decisions without human input.

Guterres reminded the assembled journalists that governments have a “first responsibility” to protect their citizens from the unintended consequences of AI. “National action is essential,” he said, “but it is not enough on its own.” In other words, the fight against a potentially hazardous technology needs both home‑grown policies and an international safety net.

One concrete idea floated in the U.N. briefing was a proposal from Dario Amodei, chief executive of the AI‑research firm Anthropic. Amodei has suggested that the world’s biggest tech players voluntarily slow down certain high‑risk projects while they hammer out common safety standards. It sounds almost idealistic – like asking rival football teams to agree on a slower pace of play – but the fact that leaders such as OpenAI’s Sam Altman and SpaceX AI’s Elon Musk have echoed the call for a pause gives it some weight.

Meta’s Mark Zuckerberg, however, nudged back a little, insisting that every AI company must “ensure its technology is safe” on its own. That sentiment underscores a persistent obstacle: profit motives and competitive pressure, especially from China, make coordinated slow‑downs a tough sell.

And the political landscape isn’t exactly smooth sailing. Former President Donald Trump, for his part, dismissed the existential AI threat as a “hoax” and accused regulators of pushing a “sick conspiracy.” Such rhetoric, Guterres warned, only deepens the divide and makes the prospect of a shared safety framework even more elusive.

Still, the U.N. secretary‑general stressed the upside of AI. He pointed to its promise in accelerating economic development, improving education, and strengthening health systems – a reminder that the technology isn’t a monolith of doom but a tool that can be wielded for good if handled responsibly.

At the upcoming high‑level session of the General Assembly, Guterres plans to press the issue further. He said the United Nations will keep “exploring practical ideas” and that the world needs a “shared understanding of how to advance the safe, secure and responsible development of AI.” In his view, when systems become truly powerful, “stronger safeguards” will be non‑negotiable.

Back in 2025, the General Assembly set up the first global scientific body on AI, a 40‑member panel that released a sobering report in July. The study warned that unchecked AI could damage mental health, become a destructive tool, and destabilise social, economic and environmental systems. It also highlighted how hard it is to keep such technology under control once it reaches a certain scale.

Guterres knows geopolitics will make the road bumpy. He urged the nations with the most advanced AI capabilities to open channels, exchange data, and agree on baseline guard‑rails. “We have to avoid this race to the bottom that could lead one day to a gigantic disaster at global level,” he said.

When asked if the General Assembly’s gathering of representatives from 193 member states would be the perfect moment for a summit on AI, Guterres admitted there wasn’t enough time to organise a formal meeting. Still, he affirmed that a “serious international summit on artificial intelligence” is essential – and that it must bring together governments, corporations, scientists and civil‑society groups.

Whether that summit materialises this year or later, the message is already out there: AI is no longer a futuristic curiosity. It is a present‑day force that can reshape economies, societies, and perhaps even the planet’s future. The United Nations is urging the world to step back, look at the bigger picture, and set up the guard‑rails needed before the technology outruns the very humans who built it.