UN partners with Google to enhance global statistical access for AI systems
The United Nations has partnered with Google to launch a new platform called the UN System Data Commons, designed to make its extensive collection of global statistics more accessible and usable by artificial intelligence (AI) systems. Built on Google’s open-source Data Commons platform, it allows users to search across data from various UN agencies using natural language queries. The system supports the Model Context Protocol (MCP), enabling AI models to directly access external datasets. A recent benchmarking study conducted by UNICEF found that major large language models performed poorly in accurately retrieving global development indicators, with an average accuracy score of just 21.2%. This highlights a growing need for reliable data sources as more users turn to generative AI tools like ChatGPT and Gemini. The platform is currently supported by $2 million from Google.org and aims to integrate nearly all UN statistical datasets onto the system by 2027, with initial support from 26 UN entities. The initiative also seeks to improve transparency in data sourcing for users of AI tools.
Tickers: $GOOGL
Topics: Artificial Intelligence Data Commons Google UNICEF United Nations