A Guardian analysis found about two-thirds of planned U.S. data centers are slated for areas that have been in drought over the past year, raising questions about site selection, permitting, and infrastructure planning.
Power availability, tax incentives, water and land have dominated data center siting decisions.
The issue is not just server cooling. Researchers say AI’s broader water footprint also runs through power generation and semiconductor fabrication, widening the implications.
Data center developers can still move forward, as some in the industry claim it is improving water efficiency through practices such as closed-loop cooling and water restoration investments.
A Guardian analysis published this week found that about two-thirds of planned U.S. data centers are slated for locations that have been in drought over the past year.
The finding arrives as the artificial intelligence buildout continues to drive major investment in nonresidential construction, particularly in large-scale data centers and the power infrastructure required to support them.
Power availability, tax incentives, water and land have dominated data center siting decisions.
For contractors, developers, and building product manufacturers, the story is no longer just about where hyperscale demand is going. It is also about whether some of the industry’s most attractive growth markets are becoming harder to entitle, cool and operate.
The data center boom has already reshaped the nonresidential map. ConstructConnect reported that total data center spending reached $2.4 billion in April 2026, pushing the year-to-date (YTD) total to $49.5 billion.
Data Center construction spending remains more than 300 percent above year-ago levels, according to the June Data Center Report, by ConstructConnect economists, Michael Guckes, and Devin Bell.
But the Guardian reporting suggests the growth may run straight into local water stress in parts of the West, South, and interior U.S.
The Guardian reported that 517 of 809 planned U.S. data centers are in areas that experienced drought over the past year.
The report also noted that large data centers can require up to 5 million gallons of water a day for cooling, or roughly the water use of up to 50,000 people. The outlet said total U.S. data center water demand could climb to 73 billion gallons annually by 2028, up from about 17 billion gallons in 2023.
Planned US Data Center construction starts value by US Census Region is shown on a map from May to December 2026. The near-term data center pipeline remains heavily concentrated in the South, which accounts for well over half of the planned spending tracked. It is important to note, however, that these projects are still in the preconstruction phase and are not guaranteed to break ground. Image and Data: ConstructConnect June 2026 Data Center Report
Cooling systems are the most visible part of the issue, but they are not the whole picture.
A University of Georgia Extension publication released June 10 said data centers need continuous cooling to prevent components from overheating and that some systems evaporate large amounts of water while others reuse a significant portion through recirculating systems. That distinction matters because it affects both permitting scrutiny and the long-run operating profile of new facilities.
Additionally, a United Nations University report released this month points to a much larger infrastructure challenge. By 2030, the report stated, the global data centers powering AI are projected to consume 945 terawatt-hours of electricity, with an associated water footprint equal to the basic annual domestic water needs of all 1.3 billion people in sub-Saharan Africa.
The UN report considered an immense sphere of impact, and criticized the narrowness of typical environmental measures, indicating that, "...AI's environmental cost is being systematically mismeasured. Most existing assessments focus on the carbon emissions associated with training large models."
The UN report added, "Yet every kilowatt-hour of electricity used to train or run an AI system also carries a water footprint, from cooling and power generation, and a land footprint, from energy infrastructure and supply chains."
The construction implications do extend beyond the building envelope. Water demand follows the AI supply chain, including power generation and semiconductor manufacturing, both of which can shape the type of projects entering the pipeline.
ConstructConnect Associate Economist Devin Bell said, “Water infrastructure demand has climbed sharply in recent years as data center construction has expanded. Water and sewage treatment construction starts rose 87.8% from 2021 to 2025, with a further 3.7% increase forecasted for 2026.”
Bell added, “Strong growth in water and sewage treatment construction will likely be needed to support these water-intensive facilities, particularly as more data centers rise in drought-prone regions where supply is already stretched.”
Data center developers can still move forward, as some in the industry claim it is improving water efficiency through practices such as closed-loop cooling and water restoration investments.
But local opposition can intensify when communities believe new projects will compete with farms, aquifers or household demand. That opposition has been reported in multiple locations in the US where data center construction has been proposed, as well as a March 2026 Gallup poll.
The polling organization found in its first poll on local data center construction that 71% of Americans oppose AI data centers in their area, with opposition centered on water, energy, pollution, and quality-of-life concerns.
That raises the odds of tougher environmental review, more aggressive disclosure requirements and greater emphasis on cooling-system design in project approvals. It could also shift advantage toward locations that can offer abundant power and a clearer water story at the same time.
Stay connected with ConstructConnect News for construction industry news and construction market analysis to stay ahead of what’s building next.
At ConstructConnect, our software solutions provide the information that construction professionals need to start every project on a solid foundation. For more than 100 years, our keen insights and market intelligence have empowered commercial firms, building product manufacturers, trade contractors, and architects to make data-driven decisions, streamline preconstruction workflows, and maximize their productivity. Our newest offerings—including our comprehensive, AI-assisted software—help our clients find, bid on, and win more projects.
ConstructConnect operates as a business unit of Roper Technologies (Nasdaq: ROP), a constituent of the Nasdaq 100, S&P 500, and Fortune 1000.
For more information, visit constructconnect.com