UN collaborates with Google to make global data accessible for AI systems

The United Nations has partnered with Google to develop the UN System Data Commons, a platform designed to make global statistics more accessible for AI systems. Built on Google’s Data Commons platform, the system allows users to search for data from UN agencies using natural language and supports the Model Context Protocol (MCP), enabling AI systems to connect directly to data sources. A UNICEF study revealed that large language models struggle with accuracy when retrieving global development indicators, with an average accuracy score of 21.2%. The study, which tested models like GPT-4o, Claude Sonnet 4.5, and Gemini 2.5 Flash, found that many responses lacked usable numbers and that models often provided inconsistent results. The UN has committed 26 entities to the Data Commons, with nearly 20 datasets available at launch, aiming to host 80% of its statistical data by 2027. Google provided $2 million in funding and technical support, with the platform hosted on a UN-governed instance. Google’s Prem Ramaswami emphasized the platform’s potential to enhance data accessibility and AI readiness.