The United Nations has partnered with Google to launch a new platform called the UN System Data Commons, aimed at making global statistics more accessible to artificial intelligence (AI) systems. This platform is built on Google’s open-source Data Commons, which organizes data from various sources into a unified format. Unlike the previous UNData portal, which required users to navigate a traditional database, the new system allows users to search for data using natural language—like asking a question in plain English. The goal is to streamline access to data from multiple UN agencies, making it easier for AI to use this information effectively. The platform supports the Model Context Protocol (MCP), a standard that allows AI systems to directly access and use data from external sources. A recent UNICEF study, which is still being finalized for publication, tested how well six major AI models could answer questions about global development indicators. The results showed that the models had an average accuracy of 21.2%, with many responses either avoiding specific numbers or providing inconsistent answers when the same questions were asked again. This highlights the current challenges in ensuring that AI can reliably and consistently retrieve accurate data. UNICEF has observed a significant rise in traffic to its data website, with more than 6 million visits per month. A large portion of this traffic comes from users clicking on links provided by AI assistants like ChatGPT. In the first nine months of this year, visits from these AI-driven links increased by 67% compared to the same period last year. While AI assistants now account for about 10% of all visits to the site, UNICEF is preparing to share the methodology, code, and data behind its study to encourage further research and transparency. The UN System Data Commons is being developed with support from Google.org, which provided $2 million in funding and technical assistance. The platform is hosted on a UN-controlled system, and the UN aims to eventually manage it independently. Currently, 26 UN entities have committed to the project, with nearly 20 of them already contributing data. The UN plans to migrate 80% of its statistical datasets to the platform by 2027, making the data more accessible and "AI-ready," according to Shantanu Mukherjee, acting director of the UN Statistics Division. Google first introduced Data Commons in 2018 to help organize public datasets into a single framework. Last year, the platform added support for MCP, enabling AI systems to query data directly. The UN’s version of the platform also tracks the source of each statistic, allowing users to trace information back to its original UN source. This is especially important as more people rely on AI tools to find and interpret data. In a demonstration, Google showed how an AI system connected to the UN data could automatically generate dashboards, charts, and written analysis by combining multiple statistics. However, Google’s Prem Ramaswami emphasized that while AI can provide useful insights, human review is essential before relying on AI-generated conclusions for important decisions.