We often think of climate change as something solely related to CO2 emissions and increasing temperatures. What about when we realize that climate change is actually a lot more than that? As AI continues to integrate itself into our daily lives, we forget that the culprit of climate change is much closer than we think. AI has not only impacted the ways in which we communicate and learn, but has also begun showing us the ways in which it can affect the future of our world’s water supply and climate.
AI data centers depend on physical thermodynamic systems that must dissipate heat generated during computation (Saleh 2026). Through this process, they become closely coupled to hydrologic systems and highly depend on and influence water and the environment. Water demand has emerged as an important consideration in AI infrastructure planning and development and may no longer be sufficient for evaluating the sustainability of AI infrastructure (Saleh 2026). This originates from one key issue, heat. AI data centers operate continuously and at high computational intensity, requiring uninterrupted cooling to maintain stable operation. Recent estimates say that U.S data centers consume approximately 449 million gallons of water per day while future AI expansion may require an additional 697 to 1451 million gallons per day by 2030 (Saleh 2026). Individual facilities can already consume several million gallons of water per day, with Google reporting annual withdrawals approaching 1.4 billion gallons at a single campus in council bluffs Iowa. Circular water systems either developed from scratch or retrofitted to facilitate water reuse are emerging as one of the key innovations in AI data center operation and maintenance (Saleh 2026). These systems are designed to drastically reduce, reuse, or eliminate the consumption of fresh drinking water due to its high demand to cool high-density AI servers.
This is especially concerning for those living close to AI data centers, as there have been many water insecurity concerns at household and community-scales associated with datacenter development (Shah 2026). Aside from affecting us directly by limiting accessible, clean drinking water, rising utility costs could be a foreseeable issue. Thus, legal filings and investigative reports reveal concerns related to not only hydrological sustainability but also rising utility costs (Shah 2026).
All in all, AI data centers use up natural resources like land, power, and water in ways most communities have never experienced (LaGrandeur 2026). This industry continues to rapidly grow, getting closer and closer to our communities.
References
LaGrandeur, Kevin. 2026. “AI Data Centers’ Use of Natural Resources.” AI and Ethics Handbook, August 15, 383–400. https://doi.org/10.1007/978-3-032-12201-8_20.
Saleh, Navid B. 2026. “Why Water Demand Is No Longer Enough: Rethinking Sustainability in AI Data Centers.” Environ. Sci. Technol. 60 (26): 18383–86. https://doi.org/10.1021/acs.est.6c08380.
Shah, Sameer H. 2026. “Four Water Insecurity Concerns About Datacenters Driving the AI Revolution.” PLOS Water 5 (1): e0000500. https://doi.org/10.1371/journal.pwat.0000500.
Leave a Reply
You must be logged in to post a comment.