Google has upgraded Earth AI, integrating Geospatial Reasoning powered by its Gemini models to help organisations better understand and respond to environmental and disaster challenges.

This upgrade builds on decades of geospatial modelling. It enables AI to automatically connect multiple datasets including satellite imagery, weather forecasts, population maps, and environmental data to deliver a more comprehensive understanding of risks and vulnerabilities.

By linking these datasets, Earth AI can provide actionable insights that previously required complex analytics and years of research. The system is already aiding real-world efforts: during the 2025 California wildfires, Google Maps deployed AI-driven crisis alerts, helping guide 15 million people to safety.

Nonprofits like GiveDirectly use Earth AI to identify the communities most vulnerable to floods, enabling targeted and timely aid delivery.

The upgrade also enhances Google Earth’s capabilities. With Gemini-powered features, users can quickly detect changes in rivers, monitor harmful algae blooms, track deforestation, and assess environmental risks in near real time.

These tools allow analysts to spot patterns and potential hazards that could affect public health, water supply, or local ecosystems. These features will initially roll out to US-based Google Earth Professional and Google AI Pro subscribers.

Additionally, Earth AI models are now accessible via Google Cloud for Trusted Testers, allowing organisations to combine their own data with Earth AI’s Population, Environment, and Imagery models.

Current pilots include WHO Africa tracking cholera outbreaks, satellite companies analysing deforestation trends, and Alphabet’s X-backed Bellwether improving hurricane prediction models for insurers.