Artificial intelligence has single-handedly changed the world as we know it. It has been embedded into society and continues to advance in complexity. Although the benefits of using AI are endless, the drawbacks have become rapidly more concerning.
AI models require significant resources, including water and energy, to function. As AI technology advances and its use expands, the demand for these resources will consequently increase.
Water is involved in the production, operation and maintenance of AI models. This includes water used to cool data centers, manufacture computer chips and generate the electricity that powers these facilities.
According to a 2025 study published in the Journal of Cleaner Production, “Recent studies show that large-scale AI data centres already consume millions of litres of water per day.” The study projects that, under a business-as-usual scenario, global water consumption by AI data centers could increase nearly sevenfold between 2030 and 2050.
Although these are staggering amounts of water, the water footprint of AI is relatively small compared to agriculture and energy production. But that doesn’t mean that AI’s growing water demand couldn’t develop into an environmental disaster with the expansion and incorporation of AI.
The concern is not just how much water AI uses, but where that water comes from. According to the 2025 study, “Many of these centers are located in regions with limited freshwater availability, including parts of Latin America, the southwestern United States, India and Australia.”
In a 2023 article published in the Partners Universal International Innovation Journal, A. Shaji George and his coauthors explain that the withdrawal of large amounts of water from rivers and streams can disrupt natural water movement and reduce the availability of water for other uses, such as agriculture and drinking water.
These concerns also extend to areas close to home. The 2025 study identifies arid regions, including the southwestern United States, as poorly suited for water-intensive cooling, while Phoenix has emerged as one of the nation’s largest AI infrastructure clusters.
A location being profitable does not make it responsible. Communities, here and elsewhere, should not have to compete for limited water supplies to support the expansion of technology that we once thrived without.
Computer science teacher Leon Tynes agrees that AI’s advancement should not come at the expense of the environment. “We cannot let the code and the machines develop so far and so fast that we don’t consider what it’s doing to animals, plants, humans,” Tynes said.
What concerns me is how quickly AI is advancing while questions about its environmental impact remain. The potential harm to our water supplies and ecosystems deserves more attention and prevention as we look ahead to the future.
“We cannot wait until we deplete natural resources to figure out that AI data centers need to be retooled and redesigned to consider the environment, the immediate and the larger environment around it,” Tynes said.
Water supports far more than the technology we create. The needs of people, plants and animals should take priority over making AI faster and more powerful.
Protecting those resources should begin with AI companies publicly reporting their water use and considering local water availability before building data centers. Communities deserve to understand the potential environmental costs of facilities being built around them, and it is often difficult to measure AI’s water usage due to the lack of transparency from company executives.
The Partners Universal International Innovation Journal states, “The recycling and reuse of wastewater generated by data centers can reduce the environmental impacts of AI systems.” Companies should prioritize ethical techniques and invest in cooling methods that require less water as these companies continue to expand.
Artificial intelligence is reshaping industries, economies and daily life, providing people with new ways to learn, solve problems and complete tasks more efficiently. However, its advancement should not come at the expense of the environment that sustains us.
