Google AI Model Tops Flu Prediction Rankings for CDC
The Centers for Disease Control (CDC) announced that a flu forecasting model developed using Google AI performed the best at predicting flu-related hospital admissions for the 2025-26 season.
This analysis, conducted by the CDC's FluSight program, evaluated 39 eligible models that predicted U.S. hospital demand over several months. The superior performance of the Google model validates the potential of combining AI technology with human expertise to improve global disease forecasting.
Google developed its predictions using a specialized AI tool called Empirical Research Assistance (ERA), which generates complex optimization algorithms across various scientific fields. This successful application highlights AI's growing role in critical public health planning.
Why it matters
- —It validates the use of advanced AI (like ERA) in predicting complex public health crises.
- —The successful model can help health organizations anticipate resource needs (like hospital beds) during flu seasons.
- —It demonstrates a powerful synergy between AI computation and scientific research.
Key facts
- Google's AI model was the top performer in predicting flu hospitalizations for the 2025-26 season.
- The CDC's FluSight program analyzed 39 competing forecasting models.
- The AI tool used, Empirical Research Assistance (ERA), generates optimization algorithms.
- The findings confirm AI's potential to improve global disease forecasting.
The full text is in the original source. Here we provide a brief summary and key facts.