{"id":1412,"date":"2026-10-13T06:09:19","date_gmt":"2026-10-13T06:09:19","guid":{"rendered":"https:\/\/amercanfirstfinance.com\/?p=1412"},"modified":"2026-10-13T06:09:19","modified_gmt":"2026-10-13T06:09:19","slug":"the-hidden-water-crisis-why-ais-indirect-consumption-outpaces-onsite-conservation","status":"publish","type":"post","link":"https:\/\/amercanfirstfinance.com\/?p=1412","title":{"rendered":"The Hidden Water Crisis: Why AI\u2019s Indirect Consumption Outpaces Onsite Conservation"},"content":{"rendered":"

Data center developers have increasingly championed innovative technologies designed to slash the amount of water needed to cool hot-running servers, particularly in drought-stressed regions like Arizona. Yet, these localized conservation efforts are failing to address the vast majority of the water consumed by the massive global artificial intelligence buildout. <\/p>\n

Miles away from the sprawling, multiplying data centers that dot the modern landscape, traditional and alternative power plants are now consuming enormous volumes of water to generate the electricity required to run artificial intelligence systems. This includes powering those very same energy-intensive cooling technologies that promise to save water at the server site.<\/p>\n

New research indicates that such indirect water consumption is rapidly growing, creating an invisible resource drain that policy-makers and local communities are only beginning to comprehend. In a recent academic paper published by University of Michigan researchers, scientists estimated that water used offsite to power three large AI data centers planned specifically for the state would boost total annual consumption by a staggering 14 billion gallons. <\/p>\n

"There has been much less attention to this indirect water usage, even though that oftentimes is significantly higher than the direct water usage and has impacts on watersheds and regional water resources," said Shaolei Ren, a professor at the University of California at Riverside who closely studies data center water consumption.<\/p>\n

The scale of this discrepancy becomes starkly apparent when looking at corporate disclosures. At Meta Platforms Inc., one of the only major technology developers to publicly divulge its indirect water consumption, such offsite use accounted for about 96% of the 21 billion gallons, or roughly 80 billion liters, of water the company consumed in 2024. This represented nearly a 120% increase from 2021 levels, according to Meta’s latest corporate sustainability report. <\/p>\n

The remaining 4% of Meta’s water footprint was consumed directly on site, mostly at data centers that also account for nearly all of the company’s massive electricity consumption. "Transparency is a key pillar of Meta\u2019s water stewardship program," a company spokesperson said in an email statement addressing the figures.<\/p>\n

Computing-intensive AI data centers can consume up to 10 times as much electricity as conventional computing facilities, with the vast majority of that energy needed simply to power and cool advanced servers. To generate that electricity at scale, fossil-fueled and nuclear power plants require large amounts of water to produce the steam needed to run turbines, linking every digital query and model training session directly to local water reservoirs.<\/p>\n

Across the United States, a growing public backlash has emphasized deep-seated concerns about local environmental impacts, strained municipal resources, and rapidly rising residential electricity bills. This mounting grassroots and municipal opposition has prompted local governments to impose moratoriums and other regulatory restrictions. According to research firm Data Center Watch, at least 45 projects valued at nearly $68 billion were blocked or delayed between April and June alone.<\/p>\n

There is a compounding risk that data centers’ rapid, unbridled growth will severely intensify the impact of climate-driven droughts and persistent heat waves, which are already stressing water supplies across wide swaths of the country. <\/p>\n

"Demand for data centers is just skyrocketing and the truth of the matter is that the power and water facilities won\u2019t be able to keep up," warned Kerri Hickenbottom, an associate professor of engineering at the University of Arizona who researches data center water use.<\/p>\n

Estimates of US data centers\u2019 annual indirect water use vary widely depending on the methodology, ranging from 211 billion gallons, or 800 billion liters, in a 2024 Lawrence Berkeley National Laboratory study, to as much as 4 trillion gallons based on a recent analysis from the environmental nonprofit Ceres. To put that higher figure into perspective, it is enough water to supply between 2 million and nearly 37 million average US households per year.<\/p>\n

\"AI\u2019s<\/figure>\n

Industry developers acknowledge the problem but point to systemic hurdles in tracking the issue. According to developers, one primary reason for the widespread lack of disclosure across the technology sector is that there is currently no standard, universally accepted protocol for calculating data centers\u2019 indirect water use. <\/p>\n

"We need industry alignment on one if we\u2019re going to collectively report indirect water consumption," said Brandon Oyer, head of energy and water for Amazon Web Services.<\/p>\n

Amazon.com Inc. reported that water efficiency at its data centers has improved by 52% since 2021, achieved in part by deploying advanced cooling systems that require less water and through novel technologies that actively reduce heat inside the facilities. "That is keeping both the onsite and the offsite water use as low as possible," said Will Hewes, Amazon Web Services\u2019 water sustainability lead.<\/p>\n

Similarly, Ben Townsend, head of infrastructure strategy and sustainability at Google, noted that the company currently does not report its indirect water use, though it has concentrated heavily on powering its data centers with renewable energy sources. "With wind, solar, storage, these projects have effectively zero water footprint," he said.<\/p>\n

Such theoretical solutions are currently being put to the rigorous test in drought-stressed data center hotspots such as Arizona, where water has become a major political and economic flash point as the Colorado River and its vital reservoirs continue to dry up. The federal government announced a 27% cut to Arizona\u2019s water allocation, with warnings that far more severe cuts are possible if environmental conditions continue to deteriorate. The state\u2019s data centers are heavily concentrated in the Phoenix metropolitan area, where local municipalities depend on the river for 20% to 50% of their total water supply.<\/p>\n

AI data center projects currently under construction in Arizona could spike onsite water demand to 5 billion gallons annually in the coming years, according to predictions by Bloomberg Intelligence. This would amount to roughly 4.5% of Phoenix\u2019s current annual water production. However, researchers emphasize that the indirect water use associated with powering those facilities would be orders of magnitude higher.<\/p>\n

Michael Gertzman, a spokesperson for the Phoenix Water Services Department, explained that a large data center applicant must legally disclose its estimated onsite water consumption during the permitting process. Crucially, indirect water use is not currently part of the evaluation.<\/p>\n

With the American Southwest suffering from a record 26-year drought, and Phoenix\u2019s summer temperatures hitting an average daily high of 108 degrees Fahrenheit, or 42 degrees Celsius, in recent years, data center developers are increasingly turning to technology that relies on little or no onsite water, such as air-chilled coolers. There are significant operational tradeoffs, however, because removing that heat from the servers requires substantially more electricity.<\/p>\n

A 2025 study co-authored by Hickenbottom noted that in the Phoenix area, the indirect water consumption of systems using air-cooled chillers can actually be more than 2.5 times that of traditional water-cooled chillers, which consume more water directly on site.<\/p>\n

Microsoft Corp., which operates several major data centers in Arizona, did not provide specific data regarding its indirect water use or indicate whether it formally calculates that consumption. The company stated that it has reduced the amount of water needed onsite per kilowatt-hour used at its data centers by 90% since the early 2000s, and it expects to save 33 million gallons annually at each of its facilities thanks to new liquid cooling designs.<\/p>\n

"Our focus is on minimizing the water we use directly in the communities where we operate while maintaining highly efficient data center designs and working with utilities to help meet our power needs responsibly," a Microsoft spokesperson said in an email statement.<\/p>\n","protected":false},"excerpt":{"rendered":"

Data center developers have increasingly championed innovative technologies designed to slash the amount of water needed to cool hot-running servers, particularly in drought-stressed regions like Arizona. Yet, these localized conservation efforts are failing to address the vast majority of the water consumed by the massive global artificial intelligence buildout. Miles away from the sprawling, multiplying […]<\/p>\n","protected":false},"author":3,"featured_media":1411,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[161],"tags":[1981,1978,164,1191,573,1977,162,163,1980,1979,165,1976],"class_list":["post-1412","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-insurtech-and-insurance-industry","tag-conservation","tag-consumption","tag-coverage","tag-crisis","tag-hidden","tag-indirect","tag-insurance","tag-insurtech","tag-onsite","tag-outpaces","tag-risk","tag-water"],"_links":{"self":[{"href":"https:\/\/amercanfirstfinance.com\/index.php?rest_route=\/wp\/v2\/posts\/1412","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/amercanfirstfinance.com\/index.php?rest_route=\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/amercanfirstfinance.com\/index.php?rest_route=\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/amercanfirstfinance.com\/index.php?rest_route=\/wp\/v2\/users\/3"}],"replies":[{"embeddable":true,"href":"https:\/\/amercanfirstfinance.com\/index.php?rest_route=%2Fwp%2Fv2%2Fcomments&post=1412"}],"version-history":[{"count":0,"href":"https:\/\/amercanfirstfinance.com\/index.php?rest_route=\/wp\/v2\/posts\/1412\/revisions"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/amercanfirstfinance.com\/index.php?rest_route=\/wp\/v2\/media\/1411"}],"wp:attachment":[{"href":"https:\/\/amercanfirstfinance.com\/index.php?rest_route=%2Fwp%2Fv2%2Fmedia&parent=1412"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/amercanfirstfinance.com\/index.php?rest_route=%2Fwp%2Fv2%2Fcategories&post=1412"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/amercanfirstfinance.com\/index.php?rest_route=%2Fwp%2Fv2%2Ftags&post=1412"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}