Environmental cost of artificial intelligence: carbon, water and soil footprints. United Nations Institute for Water, Environment and Health (UNU-INWEH).

The expansion of artificial intelligence increases the demand for data centers, chips, cooling, and power grids. Energy consumption generates impacts on carbon, water and land use. The environmental footprint depends on the amount of electricity and its origin. The effects are spread across global infrastructures and everyday uses of AI models. The environmental impact is unevenly distributed between regions and communities.