UN and Google Build AI-Ready Data Commons for Global Statistics

The United Nations is partnering with Google to make its vast collection of global statistics easier for people and AI systems to search, verify, and use.

The new UN System Data Commons is built on Google’s open-source Data Commons platform. It replaces the more traditional UNData portal with natural-language search across statistics held by UN agencies, while also allowing AI systems to connect through the Model Context Protocol, or MCP.

The initiative follows a UNICEF benchmark of six large language models and more than 133,000 responses about global development indicators. The models produced an average accuracy rate of just 21.2%. About three in five responses did not provide a usable number, and answers that were repeated on the same model versions matched only about half the time.

The underlying problem is not a lack of data. UN agencies hold extensive information on health, education, economics, demographics, and other measures, but the material is spread across fragmented databases and formats that AI systems can struggle to retrieve accurately. The new platform is designed to make each statistic traceable to its original UN source.

Twenty-six UN entities have committed to the project, with data from nearly 20 agencies available at launch. The UN aims to bring 80% of its statistical datasets onto the platform by 2027. Google.org is providing $2 million in capacity-building funding and technical support, while the system is hosted on a UN-governed instance intended to be operated independently by the organization.

Google demonstrated how an AI system connected through MCP could combine multiple indicators into dashboards, charts, and written analysis. UN and Google officials cautioned, however, that authoritative source data does not automatically make an AI-generated conclusion authoritative. Human review will still be required before outputs are cited or published.