Artificial Intelligence, powered by data centers, enhances the lives of everyday American citizens in numerous ways every day. Some of these benefits include accelerating everyday task completion and administrative workflows and enabling enhanced medical diagnostics, predictive searches, automated fraud detection, and real-time language translation. However, the bipartisan backlash against the rapid acceleration of data center construction continues to gain momentum as a host of detrimental economic, resource, and environmental impacts has come to light.
In fact, a March Gallup poll reported that 7 in 10 Americans surveyed (71%) are against the construction of AI data centers in their local area, with 48% strongly opposed. Noted reasons for their disapproval mainly focus on environmental issues but also include escalating local utility bills, excess water usage for cooling, grid instability, loss of jobs, and noise pollution. Public opinion has evolved into community action. According to Data Center Watch, a research group that tracks grassroots opposition to data center development across the United States, their study “found that data center opponents blocked or delayed at least 75 projects nationwide worth about $130 billion from January through March.”
The CRE Perspective
Reports by the various national brokerages peg the vacancy rate for data centers at approximately 1%, despite record construction activity. Much of the vacancy was driven down by hyperscalers (large-scale cloud computing providers) like AWS (Amazon), Microsoft Azure, Google Cloud, and Oracle, which operate mammoth, globally distributed data centers. FacilitiesNet reported that the 2025 inventory level reached 39 gigawatts (GW) and that 92% of the 35 GW under construction was either an owner-occupied development or pre-leased. “With supply constrained, rents increased nine percent year over year in 2025 and are up roughly 60 percent since 2020,” the report noted.
Estimates vary as to how many data centers are operating or planned in the U.S., but Data Center Map currently lists 4,677 total data centers across the 50 states. Of those, approximately 3,000 facilities are currently in operation, with over 1,500 additional in development, according to analysis by Pew Research Center. States with the largest number of data centers in operation or in various stages of planning are Virginia (685), Texas (466), and California (331). There are roughly 100 facilities in operation in both the Bay Area and Southern California, including 59 in Los Angeles, 15 in San Diego, and 14 in Irvine.
The Public Cost of Data Centers
The amount of electrical power needed to meet this intensified demand is massive. An AI-focused data center can consume as much electricity as 100,000 households. The largest ones under construction today will consume 20 times as much. According to a press release by the Electric Power Research Institute (EPRI), 4%-5% of the U.S. electricity generated is currently consumed by data centers and will more than likely double to 9%-17% by 2030. Seven states—Arizona, Indiana, Iowa, Nebraska, Nevada, Oregon, and Wyoming—may even see data centers account for more than 20% of their electricity usage.
Many data centers also require a huge amount of water for cooling. In mid-2025, the Environmental and Energy Study Institute (EESI) noted that water consumption by larger and new AI-focused data centers was increasing along with energy usage and carbon emissions. These “large data centers can consume up to 5 million gallons per day, equivalent to the water use of a town populated by 10,000 to 50,000 people,” according to EESI. Alternate cooling technologies, such as direct-to-chip cooling and immersion cooling, which require less water and are cost effective and highly energy efficient, have been implemented by data center operators. However, these may also have negative environmental impacts.
Communities also have concerns about increased carbon emissions. In many regions, developers are quoted timelines of three to seven years for grid connection, requiring them to look for alternate locations or immediate on-site power solutions. According to Utility Dive, this has created a backlog of heavy-duty combined-cycle gas turbines of up to five years. Many developers have turned to using diesel generators, raising serious local air pollution and health concerns in major data center hubs. Additionally, residents living near data centers report constant noise from cooling systems.
Communities Push Back
Despite the federal government making data center construction a national priority, there has been a significant pushback against data centers from both red and blue states. An executive order by President Trump aimed at accelerating federal AI data center policy by streamlining the permit and environmental review process has had little impact. “These directives don’t override state authority on land use, zoning, or utility regulations,” according to an April article by MultiState. At that time, 27 states were considering legislation to require “large load” customers to cover new energy infrastructure costs and report energy usage. California, Ohio, and Utah have already enacted laws that go beyond the federal government’s voluntary Ratepayer Protection Pledge. In mid-July, New York implemented a one-year statewide halt on the construction of hyperscale data centers (facilities using 50 megawatts of power or more), becoming the first U.S. state to impose a data center moratorium. Closer to home, Monterey Park became the first city in the United States to prohibit data centers with 83% of the residents voting to approve Measure NDC.
(It should be noted that the Trump administration recently approved the first-ever commercial AI data center on public land. “The Bureau of Land Management (BLM) approved an amended right-of-way that allows the Townsite Data Center project to move forward on 81 acres of federal land near Boulder City, Nevada,” as reported by Newsweek. The decision is being appealed by several groups, including Boulder City.)
Although data centers can present problems for communities, they also play a huge role in our personal lives and the growth of the national economy. The Atlantic Council makes a great case for not throwing the baby out with the bathwater in the recent “Backlash against data centers could cost the US its AI edge” essay by Jeannie Salo.
The following is an excerpt from her essay:
“The stakes of the AI race go far beyond tech
Tomorrow’s jobs, capital flows, military readiness, and economy will depend on the digital and energy infrastructure the United States builds today. This means that the United States may not lose the artificial intelligence (AI) race—and the critical advantages it confers—because its engineers fall behind China’s. It could lose because it fails to build the infrastructure AI requires—and the stakes would be hard to overstate.…
Governments and tech companies worldwide are racing to build faster—just as they once competed to build railroads, highways, and the electric grid. They should do so purposefully so that the short- and long-term benefits are both apparent and available to all citizens, as these are not niche technology debates but decisions that will shape the economic viability of every community. And they will determine the future of AI and who leads it.”
She goes on to suggest that it will take a transformative approach by all parties to come up with solutions that will benefit everyone. “Keeping its edge requires an infrastructure-intensive industrial transformation comprising new data centers, new electricity generation, semiconductors, fiber networks, cooling technologies, cloud architecture, cybersecurity, and a smarter, modern grid to power it all…Policymakers and tech companies…need to come together as a public-private ecosystem to strengthen understanding and trust alongside AI infrastructure so that AI can serve communities and their long-term interests.”
While surging demand driven by AI is fueling accelerated growth in data center development, escalating public pushback shows that local land-use and power constraints cannot be ignored by the industry. The future of AI infrastructure depends on a collaborative approach that bridges technological demands with community trust.