As climate change intensifies droughts and water stress around the world, a difficult ethical question is emerging. Who should get access to water when communities, agriculture, ecosystems, and AI infrastructure are all competing for the same finite supply?

Water has always been allocated based on a mix of law, economics, and social norms. Historically, cities were prioritized for drinking water, agriculture for food production, and ecosystems often received what remained. Data centers now enter this equation as a new and rapidly growing demand. Unlike farms or households, data centers do not directly sustain human life, yet they underpin modern economies, healthcare systems, communication networks, and scientific research.This tension makes water allocation in the age of AI a political issue as muchas a technical one.

While technology companies often emphasize efficiency gains, total water use continues to rise as AI workloads expand. Efficiency alone does not resolve the ethical challenge if overall consumption keeps increasing.

Data centers have a substantial and rapidly growing water footprint, driven primarily by the cooling demands of high-density computing and AI workloads.According to the Water Sustainability in Data Centers white paper published by Verelto,[JS1]  a relatively small 15 MW facility can directly consume between 80 million and 130 million gallons of water annually, equivalent to the yearly water use of three hospitals or two18-hole golf courses and a 100 MW data center can use about 1.1 million gallons per day (≈400 million+ gallons per year) for on-site cooling alone. On-site water consumption is dominated by evaporative cooling systems, where 70–80% of water entering a cooling tower is lost to evaporation rather than returned tot he watershed. In the United States, direct water usage by data centers has more than tripled over the past decade, rising from 5.9 billion gallons in 2014 to17.4 billion gallons in 2023, with hyperscale and colocation facilities accounting for 84% of that total.

More over, the report highlights that 75–83% of a data center’s total water foot print comes from indirect water withdrawals tied to electricity generation, underscoring that most water impacts occur “behind the meter” in power production rather than just at the facility level.

Effects of Data Centers on Communities,Farms, & Ecosystems

Communities are also impacted as they are often the first to feel the consequences. During droughts, residents may face water restrictions, higher utility costs, or declining groundwater levels. When nearby data centers continue operating without visible constraints, public trust can erode. The ethical concern is not simply about volume, but about fairness and transparency. People want to know why they can’t take a shower while industrial cooling systems continue to evaporate millions of gallons.

Agriculture adds another layer of complexity. Farming remains the largest user offre shwater globally and is essential for food security. Farmers already face shrinking water allocations, unpredictable weather, and rising costs. When water markets or permits shift toward industrial users like data centers, rural economies can be disadvantaged, particularly when fresh water availability was already unstable.

Ecosystems are often the silent stakeholder in this debate. Rivers, wetlands, and aquifers require minimum flows to remain healthy. When withdrawals increase, ecosystems degrade, biodiversity declines, and long term water resilience weakens for everyone.Ethical water allocation must consider not only human users, but also the natural systems that make water availability possible in the first place. A2023 report by WWF, “the High Cost of Cheap Water” values the direct value offresh water to be $7.5 trillion annually while the indirect value is closer to $50trillion annually. Ignoring ecosystems may provide short term gains for AI infrastructure, but it creates long term losses for society. [JS2] 

How Water Usage Transparency and EthicalPlanning Play a Role

Transparency is becoming a central ethical requirement. Many data center operators do not fully disclose water usage, sources, or impacts at the local level. Communities and regulators increasingly demand clear reporting on how much water is used, when it is used, and whether alternative sources such as reclaimed or non-potable water are being employed. Ethical water allocation cannot occur without shared data and public accountability.

Another dimension involves geography. Data centers are often built where land and powe rare cheap, not where water is abundant. This can concentrate water intensive infrastructure in arid or drought prone regions. Ethical planning would alignAI infrastructure growth with regional water realities rather than forcing stressed watersheds to absorb additional demand.

The Future of Water Scarcity in an AI World

Ultimately, the ethics of water allocation in the age of AI require a shift in thinking.Water is no longer just a utility input. It is a shared, finite resource that connects communities, food systems, ecosystems, and digital infrastructure.Decisions about who gets water reflect societal values about what matters most.As AI continues to expand, those decisions must be made deliberately, transparently, and with long term resilience in mind.

The question is not whether AI deserves water. The question is how much, from where, and under what conditions. Answering that question ethically will define whether the digital future strengthens society or deepens existing resource inequalities.

In2026, a UNreport titled [JS3] World enters eraof ‘global water bankruptcy saysthat an ‘Era of ‘global water bankruptcy’ ishere. OceanWell was established to supply more liquidity into the system,without asking the ocean to pay the price.

UN/ Global Water Reports

1. UN World Water DevelopmentReport 2025 –Flagship UN-Water analysis of global freshwater issues and water resourcetrends
https://www.unwater.org/publications/un-world-water-development-report-2025 UN-Water

2. UN. (2026, January). [World entersera of ‘global water bankruptcy]. UN News. https://news.un.org/en/story/2026/01/1166800

3. UN World Water DevelopmentReport (General Overview) –Official portal for annual WWDR reports on water and sanitation
https://www.unwater.org/publications/un-world-water-development-report UN-Water

4. UN World Water DevelopmentReport 2024 –Focus on water for prosperity and peace
https://www.unwater.org/publications/un-world-water-development-report-2024 UN-Water

WaterGovernance & Allocation Policy

5. OECD – Water Policies andAllocation –Describes water allocation regimes and guidance for equitable rules
https://www.oecd.org/en/topics/sub-issues/water.html OECD

6. OECD – Water WithdrawalsIndicator –Statistics on freshwater withdrawals from various sectors
https://www.oecd.org/en/data/indicators/water-withdrawals.html OECD

7. OECD Report: Measuring theEnvironmental Impacts of AI(PDF) – Includes discussion on water and energy use in AI compute
https://www.oecd.org/content/dam/oecd/en/publications/reports/2022/11/measuring-the-environmental-impacts-of-artificial-intelligence-compute-and-applications_3dddded5/7babf571-en.pdf OECD

DataCenter & AI Water Use

8. Data Centers and WaterConsumption (EESI) –Overview of how large data centers use water and impact regions
https://www.eesi.org/articles/view/data-centers-and-water-consumption Environmental and Energy Study Institute

9. Brookings: AI, DataCenters, and Water –Policy discussion on water demand challenges from data infrastructure
https://www.brookings.edu/articles/ai-data-centers-and-water/ Brookings

10. Veralto.(2025). Water Sustainability in Data Centers [White paper]. Veralto.Retrieved from https://cloud.kapostcontent.net/pub/6362a4c4-3922-45f2-b5e4-42466c36db41/water-sustainability-in-data-centers?kui=EOtuZA07Lk9qBeqQ-X4cyw

Academic& Scientific Research

11. Research on Data CentreWater Consumption (npj Clean Water) – Peer-reviewed paper on measurement andimpacts
https://www.nature.com/articles/s41545-021-00101-w Nature

12. SDG Policy Brief: AI’sExcessive Water Consumption –Analysis connecting AI growth and water needs
https://sdgs.un.org/sites/default/files/2024-05/Gupta%2C%20et%20al._AIs%25s%20excessive%20water%20consumption.pdf SDGs

‍