UN teams up with Google to fix AI's data problem

The United Nations has a big problem: the AI systems that could help solve global challenges can't actually read its data properly.

In a recent test, UNICEF (the UN's children's organization) asked leading AI models—the smart computer systems that power tools like ChatGPT—to find important global development statistics. The results were disappointing. The AI kept getting things wrong or missing information entirely. This matters because the UN holds mountains of data about health, poverty, education, and other critical issues. If AI can't access this information reliably, it can't help governments and organizations make better decisions.

So the UN turned to Google for help. Together, they're working to reorganize and prepare UN data in a way that AI agents—think of them as digital assistants that can search through information and solve problems—can actually understand and use correctly. This isn't about creating new data. It's about making existing data "AI-ready," almost like translating it into a language that machines can reliably work with.

Why does this matter to you? Better AI access to global data could eventually help solve real-world problems faster—from disease outbreaks to humanitarian crises. But first, the machines need to be able to trust the information they're reading. That's what this partnership is trying to fix.