From skylines to farmland, where do AI data centers fit in?

Across Iowa’s landscape, between fields, farms, and right outside of towns, clusters of concrete structures have begun to loom over communities. 

These large buildings — the components of an AI data center — are home to machines that always run. They hold hundreds of thousands of precise rows and columns. They compute millions of daily requests. These groups of structures can expand out to more than 4 million square feet — about 75 football fields. They constantly hum, generators going day and night. 

The slew of buildings — and the sound and light they emit — could be mistaken for a small city on the horizon. But inside, these data centers are nearly devoid of life. Automation makes it possible for the buildings to employ a surprisingly small staff.

Which begs the question, why is Iowa so intent on inviting these sleepless cities into its towns?

Corn, Soybeans, and Software Solutions

Artificial Intelligence is everywhere. For some, it’s integrated into their homes, adjusting the thermostat and cataloging the fridge. For others, it’s acting as an office secretary, summarizing to-do lists and emails.

For AI to think, it requires computers — millions of computers — to generate responses. These computers live across the country in AI data centers. According to DataCenterMap.com, an industry-tracking website, Iowa alone has 95 data centers in ten markets, with the Des Moines area holding the majority — 69 are located in and around the capital city. 

Data centers aren’t new. The first general data center was built in 1945. But data centers weren’t specialized for AI until the 2010s. What’s the difference? Traditional data centers are for storage and website hosting. They require fewer computers, and in turn, the energy demand is lower with less heat output, only requiring fans and air cooling.

AI data centers are much more complex and often cost two to four times more than traditional data centers. AI requires more computers, power, and advanced cooling systems to manage the intense heat generated by all those humming processors. 

To build an AI data center, there must be a reliable power infrastructure, a site with low risk for natural disasters, and the flexibility to scale. According to the American Edge Project, Virginia leads the country with 663 data centers, with 595 more on the way. Texas is a distant second at 405. Iowa sits outside the Top 10, though 55 newly announced data centers might help it move up a slot. 

“I think that Iowa is one of the few spots in the country where it’s a really hotly debated topic, because we have the water and we have the energy affordability that these companies are looking for,” says Cody Smith, the director of climate initiatives at the Iowa Environmental Council.

Across the nation, companies have been purchasing large amounts of land for development, often obtaining farmland to meet the acreage needs of the technology. The flat farmland easily transforms into concrete.

Iowa has the infrastructure — AI data centers are quick to bring the money.

Community Impact

According to the 2025 AI Index Report from Stanford University Human-Centered Artificial Intelligence, private investment in AI grew to over $109 billion in 2024, compared to only $45 million spent by artificial intelligence startups in 2010. The rising investment translates directly to new economic opportunities for those who buy in. 

The data centers that house AI bring in a multitude of new jobs, primarily in construction. Due to the importance of water cooling systems in data centers, plumbers and pipefitters are also a major component of the production and upkeep of the buildings. Recent open calls for plumbers at data center job sites have higher pay scales.

“This site [in Cedar Rapids, Iowa] was offering $13 above pay scale and over $300 per diem,” says Jonathan Widger, a second-year plumbing apprentice in Bettendorf, Iowa. “This is much more money than I know of being offered in this part of the country.”

Widger did not pursue working at the data center due to personal ethical concerns.

“[AI] reduces real people in the workforce, which reduces people’s power in voicing their opinions,” Widger says. “The noise pollution from these centers can be unbearable to anyone in the surrounding areas, and water pressure and quality for the surrounding areas tend to be extremely poor.”

To combat the noise pollution and water pressure issues, AI data centers often invest money to improve existing water utility systems. But while many companies have made positive impacts on the surrounding community infrastructure, there are no laws that require them to do so. Now, there are new proposals that would require them to. According to the lobbying firm Multistate, over 300 bills have been filed in over 30 state legislatures in 2026, touching on everything from zoning requirements to rethinking data centers’ tax incentives. 

“Some places are doing what’s called ‘Bring Your Own Clean’ tariffs, requiring them to bring their own energy,” says Josh Mendelbaum, a senior attorney at Environmental Law and Policy Center. “If they’re investing in a place and going to impact the grid, they should also be investing in wind, solar, and storage to support that grid in that same location.”

For data centers to be constantly running, there is often an increase in power usage. According to a report from the Electric Power Research Institute (EPRI), data centers consumed 4% of the total energy produced in the U.S. in 2024. That number could reach as high as 17% of all U.S. energy by 2030, according to estimates by the EPRI.

Currently, in Iowa, data centers use up 11.4% of the state’s power. More power used means more energy must be made. Energy can be sustainably sourced from wind or solar, or from dirty sources such as coal. 

“We are impacted in terms of the broader energy conversation. It’s potentially impacting our ability to meet our 24/7 clean energy goal, if the data centers are leading to coal plants [running] longer than they otherwise would,” Mendelbaum says.

If companies are not mandated to make infrastructure improvements, the surrounding environment begins to bear the strain due to a lack of appropriate resources.

The Expense Beyond the Dollar


The intensive tasks required of computers and servers in AI data centers often cause computers to overheat. To ensure that the center remains at a safe temperature, water circulates through cooling towers, absorbs the heat, and then evaporates.

According to the Environmental and Energy Study Institute, each medium-sized data center can consume up to 110 million gallons of water per year. In contrast, only 20% of the water withdrawn is returned to the rivers and watersheds. Most of it evaporates.

Beyond draining the water supply, the water that does go back into Iowa’s ecosystem is dangerous.

“The main impact on water quality from data centers is the fact that when they release the water, it’s usually at a higher temperature, called thermal pollution,” Smith says. “That can be really damaging for aquatic ecosystems and also promote bacterial growth and pathogen development.”

With data centers releasing a steady stream of high-temperature water, it can result in ‘dead zones.’ No oxygen is left to breathe, and no organisms are left to live. 

“Water quality is already in a dire situation for a lot of people in the state. Energy affordability is in a dire situation,” Smith says.

Balancing the risks


The AI economic sector is growing rapidly, and investment in it won’t stop any time soon.

“However, we need strong protections. We need very transparent requirements to know what’s going on, and we need safeguards,” Smith says. “We want to be good stewards of the environment [but] we also don’t want to exclude Iowa from a future that’s growing.” 

According to the Iowa Environmental Council, Iowa adopted a state water plan in 1985, but has not since updated or developed a more comprehensive plan. Due to a shift towards extreme weather patterns, increasing water use, and pollution, that plan is now unreliable and no longer a proper safeguard.

“The state government is meant to, in essence, play referee,” Smith says. “They should make sure that if you want to move to Iowa and you want to build a data center, fundamentally, you can do that, but we shouldn’t be sacrificing our water.” 

Despite this sentiment, some companies are attempting to evade the responsibility of their impact on water. As of early March, Google was seeking to bypass the data center zoning rules put in place by Linn County, Iowa, by annexing the land for the center campus in Palo, two miles away. 

Palo is a city entirely within Linn County borders, but has no zoning restrictions specific to AI data centers. With Google bypassing the reporting and sustainability requirements, the damage to the water will still be felt in Linn County.

With the increasing private investment, AI data centers will not stop being built anytime soon. Even if tradespeople remain conflicted on taking job offers, companies will find a way to get the AI data centers up and running.

“Companies taking advantage of smaller companies is harmful to blue-collar and working-class people. A lot of times, non-union work is done by someone not being paid what they’re worth,” Widger says. “As long as these facilities are being built, I would much rather have a union hand building them.”

Looking to the future

Many legislators in Iowa have begun proposing files and bills that detail transparency requirements for existing data centers and those yet to be built. For example, House File 2690, introduced in the 2026 session, requires the industry to increase oversight of resource consumption.

“On the energy side, they’ll have to talk about their sustainability or energy efficiency practices that they’re using,” says Jordan Oster, the energy program director at Iowa Environmental Council. “And on the water side, they’ll have to talk about how they recycle and reuse water. Those are all positive steps.” 

Thinking of individual impact can become disheartening in a system that relies on the economics of larger businesses and actions in a State House and Senate. Supporting these positive steps by contacting legislators is one way to stay engaged.

“I think transparency can make sure we’re scrutinizing things correctly, but also bring reform and change,” Oster says. “We just can’t turn away from it. We need to embrace it.” 

AI is here to stay. Its impacts depend on how we decide to use it.

No responses yet

Leave a Reply

Your email address will not be published. Required fields are marked *

css.php