Google Earth AI expanding access around the globe (blog.google)

Google announced an expansion of Google Earth AI, a suite of geospatial models powered by Gemini’s advanced reasoning, designed to accelerate environmental monitoring, disaster response and public-health interventions. Building on capabilities such as flood forecasting (now covering over two billion people) and crisis alerts used during the 2025 California wildfires, Earth AI now includes a Geospatial Reasoning framework that automatically links models—weather, population maps, satellite imagery and more—to answer complex, multi-factor questions (for example, which communities and infrastructure are most vulnerable to a flood). Google is opening wider access to these tools via Google Earth, Google Cloud Trusted Tester programs, and higher-limit Gemini capabilities for Google AI Pro and Ultra subscribers in the U.S. Technically, Earth AI packages Imagery, Population and Environment models so organizations can run them alongside their own datasets (via Google Cloud and Imagery Insights), enabling rapid object detection, pattern discovery and predictive reasoning from satellite data. Practical uses already in pilot include WHO AFRO predicting cholera risk in the DRC, Planet and Airbus mapping deforestation and vegetation encroachment, and Alphabet’s Bellwether providing hurricane insights for insurance. The move aims to shorten timelines that once required complex analytics and years of research, letting cities, nonprofits and enterprises make faster operational decisions—triaging aid, issuing water-quality warnings (e.g., algae blooms), or preempting infrastructure failures.