Image Courtesy: Google

October 23, 2025 — Google has announced a major expansion of Google Earth AI, the company’s next-generation geospatial intelligence platform powered by Gemini and decades of Earth modeling. The update brings advanced Geospatial Reasoning, new Earth AI models, and broader Google Cloud integration to help enterprises, cities, and nonprofits tackle global challenges such as environmental monitoring, crisis response, and public health management.

Built on Google’s deep expertise in AI and mapping technologies, Earth AI combines geospatial data, predictive analytics, and Gemini’s multimodal reasoning to deliver insights that once required years of specialized research — now accessible within minutes.

Helping the World Respond to Crises with AI

From natural disasters to humanitarian emergencies, Google has long used geospatial AI to power real-world interventions. Its flood forecasting models now reach over two billion people, providing early warnings that have helped organizations like World Vision deliver aid faster.

During the 2025 California wildfires, Google’s systems delivered crisis alerts to more than 15 million residents across Los Angeles, offering shelter information and live updates through Search and Maps.

These capabilities now form the foundation of Google Earth AI, uniting flood, wildfire, air quality, and cyclone models into a single AI-driven framework.

“Earth AI combines our decades of Earth modeling with Gemini’s advanced reasoning to give organizations a deeper understanding of the world in minutes,” said Yossi Matias, VP of Engineering & Research and Crisis Response Lead at Google.

Geospatial Reasoning: Connecting the Dots Across the Planet

The centerpiece of today’s update is Geospatial Reasoning, a Gemini-powered framework that lets AI automatically connect diverse Earth AI models — including weather, population, and satellite data — to analyze complex real-world problems.

Instead of analyzing single data layers, Geospatial Reasoning helps identify which communities are most vulnerable, what infrastructure is at risk, and where to focus aid during disasters.

One early partner, the nonprofit GiveDirectly, is using this framework to combine flood and population density data to direct financial aid to those most in need after major flooding events.

Social impact organizations can now apply to become Trusted Testers for Geospatial Reasoning, with expanded non-commercial access through Google.org in the future.

Deeper Insights in Google Earth with Gemini

Google is also enhancing Gemini’s capabilities inside Google Earth, enabling professionals to query satellite data and surface patterns with simple natural language prompts.

For example, water management agencies can ask Gemini to “find rivers that have dried up in the past 12 months” to forecast dust storm risk, while environmental analysts can identify harmful algae blooms in real time to protect drinking water supplies.

This experimental feature will be available in the U.S. in the coming weeks for Google Earth Professional and Professional Advanced users.

Starting today, Google AI Pro and Ultra subscribers can access expanded Gemini-in-Earth capabilities with higher query limits.

Bringing Earth AI to Google Cloud

Google is extending Earth AI’s reach to Google Cloud, giving businesses and developers direct access to its Imagery, Population, and Environment models.

Through the Trusted Tester Program, organizations can integrate these models with their own datasets — enabling use cases like:

• Environmental monitoring and sustainability analytics

• Infrastructure and risk assessment

• Disaster prediction and rapid response

Developers can also use Imagery Insights to analyze satellite data alongside Google’s proprietary models to create domain-specific applications.

Earth AI in Action: Real-World Impact

Thousands of organizations are already piloting Earth AI across sectors:

• World Health Organization (WHO AFRO) is using Earth AI to forecast cholera outbreak risk in the Democratic Republic of Congo, guiding vaccination and sanitation strategies.

• Planet and Airbus leverage Earth AI to analyze massive satellite datasets — from tracking deforestation to detecting vegetation growth near power lines that could cause outages.

• Bellwether, a moonshot from Alphabet’s X, is working with McGill and Partners to deliver hurricane prediction insights that help insurers pay claims faster and accelerate rebuilding after disasters.

Toward a More Intelligent Planet

With these new updates, Google aims to make Earth AI a foundational tool for scientists, enterprises, and humanitarian organizations alike.

“We’re working to train Earth AI models to reason about the physical world as fluently as language models reason about the digital one,” said Chris Phillips, VP & GM of Google Geo.

Google plans to continue expanding access globally, strengthening collaboration with partners in public health, crisis response, environmental science, and climate resilience.

Source: Google Blog — Updates to Google Earth AI