AI's 'Killer Robots' Alarm UN

Increasingly powerful AI chips built for civilian use are now being deployed on the battlefield, where "killer robots" have already become an almost normalized reality. Against this backdrop, UN Secretary-General Antonio Guterres on Monday called for far-reaching, worldwide regulation of artificial intelligence. Speaking at the UN's first Global Dialogue on AI Governance in Geneva, he said this generation may be the last with the chance to determine the terms on which humans and machines will coexist. He argued that if AI wants to be powerful, it must be governed; if it wants to earn trust, its builders must be held accountable; and if it wants to be truly global, it must be fair.

Guterres stressed the need to make AI more accessible to the billions of people worldwide who remain excluded from the technology's benefits. He said any future global agreement must be one that earns the world's trust and places safety as the highest priority, particularly in protecting children from digital manipulation and abuse. Echoing the call, UN General Assembly President Annalena Baerbock urged collective action against the "malicious use" of AI, citing published data showing that 99 percent of deepfakes are sexual in nature and 96 percent of those target women and girls.

Another priority for global AI balance and oversight, Guterres said, should be guaranteed access to the self-learning technology for developing countries, alongside a push for all AI data centers to run on renewable energy by 2030. He said AI sits at the center of a shared future, but one in which machines can inform while humans decide and remain accountable. He first raised the call for AI regulation at the UN General Assembly in 2017, and reiterated that position at the Geneva dialogue.

In just a few years, AI has entered the mainstream and transformed economies and societies, with effects both positive and negative. The UN has been leading efforts to build international regulation and governance for AI even ahead of this shift, a continuation reflected in Monday's inaugural Global Dialogue on AI Governance in Geneva, attended by representatives from companies, researchers, technologists, and civil society discussing how to keep humanity at the center of this transformative technology. A second Global Dialogue is scheduled for New York in May. Yoshua Bengio, co-chair of the International Independent Scientific Panel on AI, said there is no sign the technology's development pace will slow soon, adding that deeply concerning tests have shown frontier AI models capable of deceiving humans and detecting when they are being tested.

Used properly and shared broadly, Guterres told delegates, AI could compress decades of development into just a few years and become "the great equalizer" of the 21st century. But that requires extensive safety testing and clear legal accountability first. He said that when countries agree on how systems are tested, risks are measured, and liability is assigned, safety spreads along with the technology — but when they fail to agree, a fragmented patchwork of rules raises costs, divides the world, and protects no one.

Guterres said any future AI governance agreement must place children's safety and wellbeing as the highest priority, calling on nations to adopt an "AI Child Safety Pledge." Under the pledge, AI developers must prove that no system reaches children without child-focused safety testing and independent oversight, that there is zero tolerance for AI generating child sexual abuse material, and that systems must pause and connect a child to real human support if they show signs of distress. He also stressed there can be no compromise on human rights, insisting machines may inform high-stakes decisions in justice, healthcare, and policing, but humans must decide and remain accountable.

Calling for greater public investment in AI, Guterres noted that private investment in AI infrastructure is estimated at $500 trillion, dwarfing public support for capacity-building in developing countries. To narrow this gap, he announced that more than 20 countries have already backed his initiative to establish a UN-supported Global Network for Exchange and Cooperation on AI Capacity Building. He also renewed calls for major AI companies to measure and publicly disclose the full environmental footprint of their technology — including carbon, water, and land use — noting that data centers already consume more electricity than most countries in the world, and by 2030 their water use could match the annual needs of 1.3 billion people across sub-Saharan Africa.

//DBTech/BR/EK//