UN Warns AI Water Use Could Match 1.3 Billion People's Needs by 2030
Updated
Updated · TIME · Jun 3
UN Warns AI Water Use Could Match 1.3 Billion People's Needs by 2030
3 articles · Updated · TIME · Jun 3
Summary
945 terawatt-hours of electricity could be consumed by AI data centers in 2030, with their water use matching the basic annual needs of 1.3 billion people, a UNU-INWEH report said.
The report argues AI's environmental burden is being misread because policymakers focus on carbon alone, even though cooling systems drive heavy water use and power infrastructure and supply chains expand land demands.
5 million gallons a day can be used by large data centers for cooling, and local strains are already visible from Ireland—where data centers took 21% of metered electricity in 2023—to drought-hit Querétaro and Uruguay.
32 countries host AI-specialized data centers, with 90% of capacity concentrated in the U.S. and China, a gap the UN says is widening digital inequality while e-waste could reach 2.5 million metric tons a year by 2030.
UN researchers urged governments and investors to require permitting, environmental reviews and community consultation that account for water and land use alongside carbon when approving AI infrastructure.
Is the race for AI supremacy creating an unavoidable global energy and water crisis?
What is the hidden environmental price of your daily AI queries?
AI’s Environmental Footprint by 2030: The UN’s Urgent Call for Multi-Metric, Global Action on Carbon, Water, Land, and E-Waste
Overview
The 2026 UN University Report warns that AI’s environmental footprint is rapidly escalating and could become critical by 2030. Every AI interaction uses finite resources—electricity, water, land, and materials—leading to carbon emissions and electronic waste. The scale of this impact depends on how AI systems are designed, how often they are used, and what tasks they perform. The current fast-paced, competitive growth in AI often overlooks sustainability, intensifying resource demands and ecological risks. Without urgent changes in design and deployment, AI’s expansion will pose long-term threats to planetary health and resource availability.