Google DeepMind Releases Global AI Forest Map
The new 10-meter resolution map provides a standardized reference to help track global forest land usage.
Updated on Oct. 31, 2026 in Geography
Live Poll
Do you trust AI-driven satellite mapping to accurately report on corporate environmental compliance?
Google DeepMind and the World Resources Institute have launched an AI-generated map titled ForTy that classifies global land area into six categories. This tool provides a high-resolution 2020 reference point to help businesses ensure compliance with international deforestation-free product regulations.
Why it matters
The project establishes a critical 2020 baseline required by the EU Regulation on Deforestation-free Products. By providing standardized data, it allows corporations to verify supply chain sustainability and adhere to strict environmental import standards.
The ForTy map utilizes the AlphaEarth Foundations model to achieve a 10-meter resolution across the globe. Researchers incorporated tree height and forest loss data to refine classifications, successfully identifying 89% of primary forests across the 8,190 sites reviewed.
The players
Google DeepMind
This subsidiary of Alphabet Inc. specializes in artificial intelligence research and deep learning technologies.
World Resources Institute
This global research organization works to manage natural resources and promote sustainable economic development.
The details
The system processes multi-satellite data using three model stages to generate pixel-level descriptions of land use. While it maps 4.1 billion hectares of forest, it currently identifies about half of all tree crops and 60% of plantations, with ongoing work needed to improve accuracy for complex agricultural regions.
Timeline
2017: Annual satellite data coverage began in Google Earth Engine.
December 31, 2020: Cutoff date established for EU deforestation-free product regulations.
July 2025: Google Deepmind introduced the AlphaEarth Foundations model.
Fall 2025: Google integrated its Earth AI platform with Gemini.
2030: The target year for the UN goal to protect 30 percent of global land.
The Big Picture
This project advances the use of foundational AI models in environmental monitoring by replacing fragmented legacy satellite analysis with a unified, high-resolution classification system. It represents a shift toward automated, scalable land-use verification that bridges gaps between remote sensing and global trade policy.
The standardized mapping data will likely streamline how businesses verify the source of wood and agricultural commodities, potentially reducing the cost of environmental compliance. In the future, the integration of open woodlands into the map will provide a more comprehensive view of biodiversity, helping to track the efficacy of global land conservation efforts.
The takeaway
The deployment of the ForTy map demonstrates the increasing capacity for AI to solve complex supply chain sustainability challenges at a global scale. Users looking to utilize this data can access the classification layers and test data under a Creative Commons CC BY 4.0 license.
Further reading
Explore more spatial analysis and land-use updates in the Geography section.
Source note: This article includes information reported by The Decoder.
Live Poll
Do you trust AI-driven satellite mapping to accurately report on corporate environmental compliance?
