Three routes to early access
On 14 October, Google Research announced expanded early access to experimental Earth AI capabilities. Partners now have three paths: deeper geospatial analysis inside Google Earth, risk assessment and scenario forecasting through a planetary prediction engine, and an experimental agent-to-agent API. The API can embed Geospatial Reasoning skills into custom workflows, including satellite-image parsing, routing and demographic embeddings. Google is also integrating Dataset Search, which queries more than 100 million public datasets, and TimesFM for nowcasting and repeated forecasting.
Geospatial data in workflows
Earth AI brings together satellite imagery, population information, infrastructure and time series. Google names disease-outbreak forecasting, clean-energy planning and infrastructure-risk assessment among its uses. The collection includes more than 70 geospatial tools, with experimental access offered to partners.