The UN says it is working with Google on the UN System Data Commons, which lets users search for statistics from across UN agencies via natural-language queries

Authoritative global statistics are now directly queryable by AI agents, bypassing the need for manual data collation.

The United Nations has launched the UN System Data Commons, a new platform developed in collaboration with Google, designed to make global statistics accessible via natural-language queries. This initiative leverages Google's open-source Data Commons platform and replaces the traditional UNData portal. The new system is engineered to support the Model Context Protocol (MCP), enabling AI systems to directly access and query the UN's data repositories. This development comes as studies, such as a UNICEF benchmark, reveal low accuracy rates (21.2% on average) in large language models when answering questions about global development indicators, often due to their inability to reliably surface authoritative data. UNICEF has observed a significant increase in traffic to its data website from generative AI assistants, accounting for approximately 10% of total visits. The UN System Data Commons aims to integrate 80% of the UN system's statistical datasets by 2027, with 26 entities currently committed and data from nearly 20 available at launch. Google provided $2 million in funding and technical support for the platform's infrastructure, with the UN intending to manage and scale it independently.

The UN System Data Commons represents a significant step towards making complex, multi-agency data accessible to AI, addressing a critical gap in the reliability and accuracy of AI-generated information on global development. The UNICEF benchmark highlights the current limitations of LLMs in handling statistical data, underscoring the need for structured, directly accessible data sources. By supporting the Model Context Protocol (MCP), the platform allows AI agents to retrieve and contextualize information, potentially improving the quality of AI-driven analysis and reports on international issues.

This initiative signals a broader trend of governments and international organizations preparing their datasets for an AI-driven future, recognizing the increasing reliance on AI tools for information retrieval. For AI developers and users, this means a more direct pathway to accurate, traceable data, reducing the risk of misinformation and improving the utility of AI assistants in research and policy analysis. The UN's goal to onboard 80% of its statistical datasets by 2027 suggests a sustained effort to integrate AI capabilities into global data governance.