---
title: "Data Centers"
description: "As data center hyperscalers continue their rapid expansion, public sentiment is growing fraught with increasing backlash and skepticism toward the vision of…"
url: https://www.independentpress.com/article/data-centers
date: 2026-08-21
categories: ["Technology","Environment","Politics"]
author: "Allie Pallotta"
---

# Data Centers

![Data Centers](https://images.ctfassets.net/ewtdlsoyixc1/3ZxnhdCbZpB7SGqp8UZf1K/21f3c7a919964caf9701e3a4cef5bf29/Allie.avif)

In the past few years, Artificial Intelligence (AI) has rapidly crept into most industries, and with it, the need for infrastructure to house and support the emerging technology. Backed by large [hyperscalers](https://datacentremagazine.com/news/top-10-hyperscalers)—including Amazon, Google, Microsoft, and Meta—as well as generous funding from investors and governments across the world, [the AI buildout has ballooned to a projected $7.6 trillion in spending on data centers and other infrastructure](https://www.goldmansachs.com/insights/articles/tracking-trillions-the-assumptions-shaping-scale-of-the-ai-build-out). The most recent estimates forecast that globally companies are on track to [spend $1 trillion in 2026](https://www.goldmansachs.com/insights/articles/global-investment-is-forecast-to-exceed-1-trillion-in-2026) on AI capital expenditures. As data centers continue to come online, the prospect of economic development, job creation, and productivity gains sounds appetizing to many. However, as more data centers are being developed, sanguine optimism is turning into worries that the AI infrastructure buildout is biting off more than it can chew. As hyperscalers continue their rapid expansion, public sentiment is growing fraught with increasing backlash and skepticism toward the vision of productivity that the world’s technological visionaries have promised. 

One of the most frequently cited concerns is AI’s seemingly insatiable thirst for water. Liquid cooling systems necessary for managing temperatures inside data centers require hundreds of millions of gallons of water every year to protect the servers that process the billions of AI queries that are received every minute. It is estimated that North American data centers consumed [1 trillion liters (or 264 billion gallons) of water in 2025](https://www.mordorintelligence.com/industry-reports/north-america-data-center-water-consumption-market). The United Nations estimates that globally data centers could consume as much as [9.3 trillion liters (2.46 trillion gallons)](https://collections.unu.edu/eserv/UNU:10647/UNU-INWEH-Report-The_Env_Cost_of_AI-2026.pdf) of water by 2030. 

In addition to the sheer volume of water consumed, where these data centers are being built and drawing water from is also consequential. Data centers built in drought-stricken areas, such as Nevada and California, can put a strain on already stressed freshwater reserves. The Western United States is presently dealing with a water and electricity crisis stemming from a more than [25-year drought in the Colorado River Basin](https://thehill.com/policy/energy-environment/6012492-colorado-river-plan-arizona-nevada-california/). The Colorado River, which supplies water for agriculture and generates hydroelectric energy via the Hoover Dam, has begun to dry up from overuse while hotter temperatures compound the problem by limiting the amount of snowmelt that typically feeds the river system. Lake Mead and Lake Powell—two artificial reservoirs linked to the Colorado River Basin—have reached historic, critically low water levels, forcing states to drastically cut back and ration their water reserves for the coming decades. 

Even in places where water is abundant, data centers are threatening water reserves. In the Great Lakes region of the northern United States, [which holds about 20% of the planet’s freshwater](https://news.wttw.com/2025/08/20/mega-data-centers-could-drain-water-supplies-great-lakes-region-if-protections-aren-t), hyperscalers have been lured by extensive groundwater reserves that can be consumed to cool their data centers. However, these same reserves are used by millions of people across Michigan, Wisconsin, Minnesota, Ohio, and Illinois for fresh drinking water. The Great Lakes’ states currently do not have groundwater management mechanisms set up to prevent the potential overextraction of these resources. While several organizations like the [Alliance for the Great Lakes](https://greatlakes.org/) seek to conserve the region’s freshwater resources, AI data centers are threatening to drain and deplete these reserves before safeguards can be put in place. 

Compounding the problem is the amount of [heat that data centers radiate](https://www.cnn.com/2026/03/30/climate/data-centers-are-having-an-underrported). A study authored by Andrea Marinoni of the University of Cambridge in March 2026, and waiting to be peer-reviewed, found that [data centers increased surface temperatures on average by 2.07 °C (or 3.73°F) and up to 9.1°C (or 16.4°F) after the start of operation](https://www.researchgate.net/publication/403073048_The_data_heat_island_effect_quantifying_the_impact_of_AI_data_centers_in_a_warming_world). For a planet already dealing with [severe drought](https://www.cbsnews.com/news/colorado-river-water-crisis-trump-cuts/), [record global temperatures](https://wmo.int/news/media-centre/new-report-suggests-more-global-temperature-records-ahead), record-breaking wildfires in the [United States](https://www.yahoo.com/news/us/articles/worst-years-us-wildfires-21st-131000898.html), [Canada](https://www.theguardian.com/world/2026/jul/29/canada-wildfire-emergency-agency), and [Europe](https://www.npr.org/2026/07/28/g-s1-135880/photos-spain-france-wildfires-climate-change), and historic loss of ice sheets in the [Arctic](https://www.pnas.org/doi/10.1073/pnas.2614134123) and [Antarctica](https://www.asoc.org/learn/antarctic-ice-and-rising-sea-levels/), these findings should be alarming. The fear is that increasing temperatures will only further exacerbate dwindling water supplies, creating a negative feedback loop that self-perpetuates. Data centers pose a significant risk of accelerating these climate-related concerns and increasing the current water crisis.

Aside from water, data centers also require considerable amounts of energy to operate. The [United Nations](https://unu.edu/inweh/news/environmental-cost-of-AIs-Enrgy-use-carbon-water-and-land-footprints)estimates that by 2030, data centers will consume 945 terawatt-hours of electricity annually, a figure higher than Japan’s current total annual energy consumption and nearly triple the combined annual energy use of Bangladesh, Nigeria, and Pakistan. If treated as a nation, the 448 terawatt-hours of electricity that global data centers consumed in 2025 would have been the world’s 11th largest electricity consumer, ahead of Saudi Arabia and behind France. According to the [Sierra Club](https://www.sierraclub.org/articles/2025/08/fools-gold-when-700-gigawatts-data-centers-come-knocking), U.S. data centers requested over 700 gigawatts (GW) of power-connection development in 2025 (more than the [entire 477 GW of electricity the U.S. consumed in 2023](https://www.eesi.org/articles/view/data-center-power-demands-are-contributing-to-higher-energy-bills)). In many locations across the country, the U.S. power grid infrastructure and transmission capacity are insufficient to keep pace with the surge demand from new data centers. To address this issue, energy utilities are attempting to upgrade and expand their transmission lines and energy infrastructure, with power grids turning to building more coal and natural gas plants to satisfy the need. A significant portion of the costs associated with surging energy demand has been passed on to consumers. Residents in Virginia, a state with a high concentration of data centers, [experienced a 13% year-over-year increase and a 30% rise in residential electricity prices from 2021 to 2025, a rate that outpaced inflation](https://www.simcorepartners.com/insights-/understanding-virginia-s-rising-electricity-prices). As a result, AI data centers are beginning to jeopardize both the efficiency of the current electric grid and the affordability of electricity that is passing through it.

Beyond the increasing strain that data centers are placing on scarce water and energy resources, they are releasing byproducts that are harmful to the environment. Data servers often emit high- and low-frequency noises from their fans and generators 24 hours a day, 7 days a week, with noise levels reaching up to 96 decibels—to put in perspective, 90 decibels would require ear protection. For neighborhoods in proximity to data centers, the constant noise pollution has been linked to medical ailments such as nausea, headaches, and even tinnitus. In addition, data center development is devouring thousands of acres of land, which can disrupt local ecosystems and raise temperatures and emissions by eliminating trees and plants in the surrounding regions. [AI data centers are also significant polluters; with estimates suggesting they discharge over 2.5 million annual tons of e-waste](https://unu.edu/inweh/news/environmental-cost-of-AIs-Enrgy-use-carbon-water-and-land-footprints), not to mention hundreds of millions of tons of greenhouse gases from the gas-guzzling generators powering their servers. Concurrently, semiconductors, which are critical hardware components that help power data centers, are also becoming significant contributors to environmental pollution. [Semiconductor manufacturing produces high global warming potential (GWP) gases](https://www.epa.gov/eps-partnership/semiconductor-industry) and [discharges toxic chemicals, including PFAS, into tens of thousands of tons of wastewater](https://pmc.ncbi.nlm.nih.gov/articles/PMC12510042/), poisoning human and non-human bodies alike. The AI data center buildout therefore not only poses an environmental threat, but also an existential risk to public health.

Environmental concerns are not the only issues that data centers have brought to the forefront of public scrutiny; fears over worsening rural-urban economic inequalities have also recently stepped into the limelight. Most data centers in the United States are being built in rural areas, particularly in the Midwest, Appalachian, and Southwest regions of the country. The state of Virginia is especially noteworthy. Virginia has seen a substantial increase in the number of data centers being built; as of August 2026, [the state hosts 670+ data centers](https://www.datacentermap.com/usa/virginia/) (roughly 14% [of the 4,700+ data centers nationwide](https://www.datacentermap.com/usa/)). Most of these data centers are being developed in rural counties with low-income communities where resources are scarce. Given the fierce resource competition and pollution caused by data centers and their operations, local communities are often bearing the brunt of these costs and deleterious side effects. While tech elites, investors, and real estate brokers within the private sector have raked in billions of dollars in profits from the AI infrastructure buildout, many consumers have been footing the costs through higher utility bills and [higher prices for consumer electronics](https://www.aol.com/articles/apple-ceo-sends-strong-warning-170700000.html?guccounter=1&guce_referrer=aHR0cHM6Ly93d3cuZ29vZ2xlLmNvbS8&guce_referrer_sig=AQAAAL-34cvIHKPO-9T1X1XlgJoMTQieY8a0MZBbyRpOleSJn8-R0GrmpOFyUPXDRlghXbTB00Vkifn6-rXa4VSVVDeZnx7cML-rjSJiaNs3DjMkog0you0NUJ_g54w86cLfcwC-XjCvlltu_n9ZXTBAba6gNxN91F9xOET1iavz9Oux). Because data centers require a vast amount of memory and storage, AI ventures are quickly eating up the semiconductors and computer memory chips that power almost all modern electronic devices. Subsequent price increase in goods—like computers and iPhones—that use these chips is putting even more pressure on consumers amid growing inflationary concerns in the United States. Continued increases in the price of electricity and electronics are undermining data centers’ utility to consumers and causing many to start questioning whether their benefits are worth the costs.

Communities across the U.S. are beginning to express frustration over public accountability and who gets a voice in major decisions regarding data center construction. American voters have increasingly voiced anger that data center projects are being approved with insufficient public input despite their potentially significant impacts on local infrastructure, public health, and consumer prices. In Box Elder County, Utah, for example, residents pushed back against [a proposal backed by billionaire investor Kevin O’Leary](https://www.sltrib.com/news/2026/07/29/oleary-box-elder-county-data/) to construct a data center almost four times the size of Central Park. Many residents criticized what they viewed as a lack of transparency and meaningful opportunities for public participation. However, rather than addressing these concerns directly, [O’Leary suggested that organized, out-of-state protesters were fueling the opposition](https://www.sltrib.com/news/2026/05/05/kevin-oleary-says-protesters/), illustrating growing tensions between local communities and powerful actors in the tech-driving expansion of AI infrastructure. Similar concerns have emerged in Memphis, Tennessee, where Elon Musk’s xAI operates the Colossus 2 supercomputing data center. To meet the facility’s enormous electricity demands, [the company has been accused of operating dozens of natural gas turbines without obtaining the necessary environmental permits, making them illegal](https://www.selc.org/news/xai-built-an-illegal-power-plant-to-power-its-data-center/). Residents of nearby neighborhoods, which are composed predominantly of black and lower-income Americans, have reported increased noise pollution and concerns about air quality, while civil rights organizations and environmental groups have argued that the emissions disproportionately burden communities that have historically faced environmental injustices. In both Box Elder County, Utah and Memphis, Tennessee, the voices of local communities have been drowned out by the profit motivations and unscrupulous business tactics of Big Tech.

There is no place where dissent and grievances toward data centers are more apparent than on the campaign trail, where the issue is quickly becoming a focal point for the 2026 U.S. midterm elections. Progressive political candidates like Abdul El-Sayed, who won the Michigan Democratic primary to run for the U.S. Senate in the upcoming midterms, and Gina Hinojosa, the Democratic nominee for Texas governor, have made data centers a centerpiece of their campaigns to appeal to voters in rural areas and other communities who have been adversely impacted. El-Sayed has advocated for more government oversight on data centers and called for AI infrastructure to serve as public-benefit corporations in [his policy platform](https://abdulforsenate.com/2026/06/dr-abdul-el-sayed-announces-ai-policy-platform-ai-under-democracy/), while [Gina Hinojosa has called for a moratorium on data center development](https://ginafortexas.com/priority/data-centers/). These policy proposals speak to a broader desire to slow the brakes on AI’s rapid advance on the daily lives of Americans. Voters’ negative views on data centers have not only offered political hopefuls a means of bridging the rural-urban divide, but also a way of overcoming partisan divides. According to a [Politico poll](https://www.politico.com/news/2026/07/21/poll-data-centers-democrats-moratorium-01001799), support for AI data centers declined among voters of both parties between January 2026 and July 2026. Among Americans who voted for Kamala Harris, support for data centers fell by thirteen percentage points while opposition increased by twenty percentage points. Trump voters likewise saw an eleven-point decline in support for data centers and an eight-point increase in opposition over the same period. Many voters are starting to unify in opposition as data centers become a political rallying cry to push back against the tech elite who are driving their development. In doing so, people are starting to reclaim power and leverage against the encroachment of AI infrastructure in their communities.

As the buildout of AI infrastructure ravenously competes for space and resources without any checks on its appetite, many Americans have been left without a voice and made to foot the bill. Like any major industry, artificial intelligence needs guardrails and thoughtful regulation to ensure that its costs do not overshadow its potential benefits. While the government and other regulatory bodies have been slow to enact checks on AI, steadily increasing grassroots opposition toward the development of AI infrastructure promises to deliver a reckoning against its immense power. As resistance toward AI data centers continues to grow, the upcoming 2026 midterm elections offer American voters a chance to have their voices finally break through the hum of AI’s insatiable appetite and ever-growing infrastructure.