Google AI model tops CDC flu hospitalization forecast for 2025-26 season

A Google AI model beat 38 competitors to become the most accurate system for forecasting flu-related hospital admissions in the 2025-26 season, according to the CDC's FluSight end-of-season analysis. The model was built with Empirical Research Assistance, a tool whose underlying technology is now being rolled out to trusted testers after publication in Nature.

• Google beat 38 competitors in flu hospitalization forecasting
• Ranking comes from CDC FluSight end-of-season analysis for 2025-26
• Model was developed using Empirical Research Assistance
• ERA technology opens to trusted testers after Nature publication