Google has been organizing the world's geospatial information for decades, making it accessible through various products such as Google Maps, Street View, and Google Earth. Geospatial information is essential in everyday situations and for a wide range of real-world enterprise problems, including public health, urban development, and climate resilience. Google's data, real-time services, and AI models can accelerate analyses and augment proprietary models and data. However, geospatial information can be big, complex, and hard to understand, requiring specialized sensors and platforms. To address these challenges, Google introduced two pre-trained, multi-purpose models: the Population Dynamics Foundation Model and a new trajectory-based mobility foundation model. These models have been tested by over 200 organizations, and Google is expanding the dataset to cover more countries. Google is also exploring how generative AI can reduce the cost, time, and domain expertise required to combine geospatial capabilities. The company is introducing new remote sensing foundation models for experimentation and a research effort called Geospatial Reasoning, which aims to bring together foundation models with generative AI to accelerate geospatial problem-solving. Geospatial Reasoning will provide rapid, trustworthy answers to complex natural language queries, and the models will be available through a trusted tester program. The initial testers of the remote sensing foundation models include WPP, Airbus, Maxar, and Planet Labs, which plan to use the models to pioneer AI-driven audience intelligence, extract insights from satellite images, and simplify and accelerate insights for their customers.