Despite decades of progress in official statistics, the people most at risk of being left behind, including women, persons with disabilities, and other marginalized groups, remain undercounted and under-served by national data systems. This gap is no longer only a question of equity. As statistical systems adopt AI, models can reflect only the data they are built on. Where inclusive data is missing, AI does not stay neutral. It reproduces and scales those blind spots, embedding bias and exclusion into the insights and decisions that follow.
The session examines how the statistical community can both measure inclusion gaps and close them by adopting new tools and approaches to data and through coordinated efforts at the global, regional, and national level. On measurement, the session will present tools such as the Open Data Inventory (ODIN) on the availability and openness of official statistics and the Inclusive Data Compass on inclusive data systems and their enabling environment. On action, the session will look beyond traditional sources and methods of analysis toward intersectional analysis, including by highlighting the work of the Steering Committee on Intersectional Analysis, to make intersecting identities more visible while strengthening, rather than bypassing, official statistics.
Speakers will connect measurement to use, showing how complete and inclusive data is the foundation for credible analysis, sound policy, and trustworthy AI that serves everyone rather than only the already-counted.