Google AI model ranked first in forecasting flu hospitalizations

A flu forecasting model developed with Google AI was the closest match to observed flu-related hospital admissions in the United States during the 2025–26 season. The Centers for Disease Control and Prevention announced the result this week.

The CDC’s end-of-season analysis compared 39 eligible models. Google’s model ranked first after forecasting hospital admissions for the current week and the next three weeks.

How the CDC compares forecasts

From October through May, the CDC’s FluSight program collects weekly forecasts from government, industry and academic teams. The combined forecast is used to communicate expected state-level demand for medical services.

Tool used to develop the model

Google’s forecasts were developed with Empirical Research Assistance, or ERA, an artificial intelligence tool that generates optimization algorithms for scientific fields. Research on ERA was recently published in Nature.

The technology underlying ERA is now available to trusted testers through Google’s experimental science tools. The CDC said the result strengthens its confidence in combining artificial intelligence with human expertise to forecast diseases.