AI firms risk 1.3 bn people's water needs by 2030, bombshell UN report says urgent action needed
AI companies like OpenAI, Google DeepMind, Microsoft AI, Anthropic, and xAI are consuming billions of gallons of water per year. Artificial intelligence (AI) is expected to use so much water by 2030 that the amount could match the basic yearly water needs of 1.3 billion people, as per a new report by the United Nations University Institute for Water, Environment and Health (UNU-INWEH). The report says people often focus only on AI’s carbon emissions, but AI also has a huge impact on water resources.
Data centers that power AI require huge amounts of water to cool servers and support their operations. Lead author Miriam Aczel said in an UNU-INWEH press release cited by TIME report, researchers were surprised that some energy options that look environmentally friendly from a carbon perspective can actually be worse for water resources. Aczel warned that focusing only on carbon emissions could create the false impression that renewable energy makes AI infrastructure completely sustainable.
Why AI needs so much water
AI uses a large amount of water mainly to keep systems cool. Powerful servers run AI models and these servers produce a lot of heat. If the machines overheat, they can slow down, get damaged, or stop working. To prevent this, data centers use water-based cooling systems like evaporative cooling towers.
These systems use water to absorb heat and keep the equipment at safe temperatures. Data centers also need water indirectly, because water is used in making electricity and building materials. Inside data centers, servers use electricity to do calculations and store data. Some of this electricity turns into heat, which must be removed to avoid overheating and system failure, as explained by the Environmental Law Institute.
Also read: US credit card debt hits $1.25 trillion as more borrowers fall behind: Are Americans heading back 18 years?
AI’s hidden water problem
The report says AI’s environmental impact is already being felt around the world, not just in future projections. Researchers warn that this level of water consumption can be dangerous for regions already struggling with drought and water shortages.
Another report by the EESI Environmental and energy study institute says the impact of data centers on water supplies is already being felt in some communities. In Northern Virginia, often called the world’s data center capital, more than 300 data centers consumed nearly 2 billion gallons of water in 2023, a 63% increase from 2019. Loudoun County alone used about 900 million gallons of water that year, leading local water authorities to rely heavily on drinking-quality water supplies to meet demand.
Data centers using more water
In Querétaro, proposed fast-tracked data center projects have raised concerns about local water supplies during ongoing drought conditions. In Uruguay, plans for a water-intensive data center sparked controversy during a severe drought in 2023. The drought reduced freshwater reserves in Uruguay’s largest city and made tap water unsafe to drink.
Residents protested because they felt that industrial projects were being given more importance than people’s access to drinking water. To stop communities from suffering because of AI growth, the report says we need a more responsible AI system. It asks governments, investors, and financial institutions to create stronger rules to reduce environmental damage caused by AI.
Researchers also say that permission for new data centers should include checks on water use and local community concerns. The report clearly warns that AI’s rising water demand could put more pressure on people and natural resources if governments do not take action soon.