A flu forecasting model developed by Google AI has been recognized by the CDC as the most accurate predictor of flu-related hospital admissions during the 2025-26 season. The CDC’s FluSight initiative collates predictions from various sources to estimate hospital admissions weekly, aiding in medical resource planning.
The recent analysis found that Google’s model was the top performer among 39 contenders, illustrating the effectiveness of combining AI with human expertise in disease forecasting.
Google’s forecasting utilized their Empirical Research Assistance (ERA) tool, which creates optimization algorithms for various scientific applications. This technology is now available for testing as part of Google’s experimental science initiatives.
