Robot Dogs, Teslas, and Rescue Helicopters: The UN AI Summit Was a Lot

Dodge past the live onstage coding sessions, AI refresher courses, an obstacle course of gizmos, round people walking round with glowing green silent-disco-style headphones blaring UN panel discussions into your ears, and you can take a pause for breath. But you might find yourself in the Networking Zone, on a rotating seating contraption called UFOTECH that looks more like the kind of lazy Susan you’d encounter at a Chinese restaurant than the networking bench it is designed to function as.

This is the AI for Good summit, organized by the United Nations’ International Telecommunication Union (ITU), where representatives from the private and public sectors try to discuss how to harness the technology for the benefit, rather than the detriment, of humanity.

While Silicon Valley execs and AI lab leaders are testifying to lawmakers in Washington about the risks of superintelligence, and the White House slaps export controls on chips, the UN AI for Good Summit—now in its 10th year—is focused on much more idealistic goals.

“Our conviction that artificial intelligence, deployed responsibly, could help solve humanity's most pressing problems—from hunger to disease to a warming planet,” Doreen Bogdan-Martin, secretary-general of the ITU, said in a keynote on the conference’s main stage. “Today, that idea is being tested, including by the challenges AI itself is bringing, even as we strive to use it for good.”

What good means—and what good it does humanity—was a question riven throughout the conference, which spread across a massive 106,000-square-meter convention center on the fringes of Geneva’s airport district. Sessions were backed by a drumbeat of worry that indifferent deployment by unchecked corporate monopolies is already hardwiring global inequality and eroding human rights.

For some on the front lines, the tech industry's utopian veneer has already worn off. Speaking on the sidelines of the event, Giulio Coppi, senior humanitarian officer at campaign group Access Now, called out the humanitarian and public sectors’ overreliance on big tech. “We should be out of the age of innocence,” Coppi says, demanding that organizations stop treating tech companies “as your best friends.” He points to a decade of opaque, multimillion-dollar deals funded by public money. “You can’t even explain what's inside your tech stack, because it has kept changing,” he warns.

Coppi’s opposition was muted compared to some: Pro-Palestine activists stormed the stage during a keynote by Amazon chief technology officer Werner Vogels, alleging that the company’s technology is being used by Israel against Palestinians, before eventually being bundled out of the venue.

“When we’re talking about AI, we love the hype, we get excited about it,” says Vijay Janapa Reddi, an engineering professor at Harvard University, over the din of competing sessions during a presentation. “The damn thing never actually lands in practice.” The problem, he says, is that “good” is too vague a standard to engineer against. “When you’re an engineer, good means nothing. I can’t build you something that is good. A plane that flies for five minutes ain’t no good.”

Much of the global debate around AI is now framed around access: Who can use the models, who can buy the chips, and who is excluded from the compute economy. It’s part of the reason why the Trump administration has implemented, then lifted, export controls on leading frontier AI models, and China is reportedly mulling making its open-weight models less open. Tightening access and cutting out poorer countries can leave them dependent on foreign infrastructure platforms and standards.

In a session on AI hardware and the widening digital divide, speakers argued that compute is no longer merely a technology problem, but a development problem. “If we mean AI for good, meaning compute for all, we must recognize that this is [about] development infrastructure, not just technology,” says Syed Munir Khasru, chairman of the Institute for Policy, Advocacy, and Governance. Others pointed out most large language models remain structured around English, making smaller, local LLMs running on cheaper hardware essential if AI is to serve communities beyond the richest markets.