A Google AI-powered model outperformed 38 competitors to become the top-performing system for forecasting flu-related hospital admissions during the 2025-26 season, according to the Centers for Disease Control and Prevention.
The CDC’s FluSight initiative aggregates weekly predictions from government, industry, and academic teams to track current and future U.S. hospital admissions, helping officials anticipate healthcare demand across states. In this season’s end-of-season analysis, Google’s model most accurately matched actual admissions, highlighting the growing impact of artificial intelligence in public health forecasting.
While there’s been negative backlash against the continued proliferation of AI use within the arts and socio-political movements, there are still some applications (such as this in particular) where advanced computing and language models can contribute to significant advancements in scientific research.
Google says that the model was developed using Empirical Research Assistance, an AI tool which is capable of generating algorithms for scientific research. Following its recent publication in the journal Nature, the technology behind ERA is now being rolled out to trusted testers through Google’s experimental science tools.
