The United Nations launched the UN System Data Commons. It is built on Google’s open source Data Commons platform. It replaces the existing UNData portal. Users can now search UN statistics with natural-language queries. The platform supports the Model Context Protocol for direct AI agent access. Twenty-six UN entities have committed to the platform. Nearly 20 are available at launch. The UN aims to include 80% of its statistical datasets by 2027. A UNICEF test found leading AI models scored 21.2% accuracy on global development statistics. Three in five model responses provided no usable number. Google.org provided $2 million in funding and technical support. The system is hosted on a UN-governed instance. It will eventually be maintained independently by the UN.
What changed
UN data was accessed via a traditional database interface on the UNData portal.
What it unlocks
Searching UN statistics with natural-language queries and direct AI agent access via MCP.
- 21.2% average accuracy in UNICEF benchmark
- 67% YoY rise in ChatGPT referrals to UNICEF data
- 6.4% of sessions from AI assistants
- 80% of UN system datasets by 2027
Sources