SerpApi.Org
If artificial intelligence had a say in where the massive data centers powering it should be built, the answer would not be “anywhere and as fast as possible.”
When Newsweek asked several leading AI models where new data centers should go, they largely agreed on a cautious approach: prioritize areas with adequate power and water resources, avoid environmentally sensitive locations, engage local communities and carefully weigh long-term impacts before breaking ground. Yet across the United States, residents say the reality often feels far different as a race to expand AI infrastructure collides with concerns over transparency, resource use and local oversight.
As tech companies and government officials push rapid expansion to meet soaring demand for AI computing power, communities from Virginia to Texas and beyond have raised concerns about electricity consumption, water use, noise and the pace of approvals.
The nationwide scale of data center concerns is evident in a crowdsourced map launched by environmental activist Erin Brockovich, which tracks facilities across the country and collects reports from communities questioning their potential environmental and quality-of-life impacts. The map suggests that while data centers are often promoted as critical infrastructure for the AI economy, debates over where and how they should be built are unfolding far beyond a handful of high-profile projects.
What Is a Data Center, and Why Are They Needed?
A data center is a facility that houses thousands of computer servers, networking equipment and storage systems that power the internet, cloud computing and artificial intelligence.
Every time someone uses a search engine, streams a video, stores files online, uses GPS or interacts with an AI chatbot, those requests are processed in a data center.
The recent boom in construction is being driven largely by AI, which requires far more computing power than traditional internet services, creating demand for larger facilities with specialized hardware and massive energy supplies.
Why Are Data Centers So Controversial?
Data centers have become a flashpoint in communities across the country because they consume large amounts of electricity and water while often being built close to residential areas.
Opponents argue the facilities can strain power grids, increase utility costs, generate noise and alter local landscapes, while some question whether the number of permanent jobs created justifies the scale of the projects.
Supporters counter that data centers bring construction jobs, tax revenue and critical digital infrastructure needed to support economic growth and maintain U.S. leadership in AI.
What Areas Are Building the Most Data Centers?
Data center development is heavily concentrated in a handful of regions with access to power, land and fiber-optic networks.
Northern Virginia remains the nation’s largest hub, while Texas and California also host a large concentration of centers.
New projects are increasingly spreading into states such as Arizona, Georgia, Ohio and Pennsylvania as developers search for additional power capacity and suitable land to support AI-related growth.
What Did AI Say About Building Data Centers?
When asked where data centers should be built, major AI systems generally favored a cautious approach that balances economic benefits with community concerns. Rather than advocating for unrestricted development, AI responses emphasized locating facilities in areas with adequate power and water resources, robust transmission infrastructure, renewable-energy potential and distance from dense residential neighborhoods. The systems also consistently highlighted the importance of community consultation, environmental safeguards, transparency and modern efficiency standards to reduce impacts on local residents.
Their responses are:
Microsoft Copilot
Microsoft Copilot, the artificial intelligence assistant developed by Microsoft, had a clear message when asked about its ideal data center construction plan.
“Data centers should go where they create the fewest burdens on existing communities—not simply wherever land is cheap or tax incentives are largest,” the AI said.
Its ideal locations for data center construction would be existing power-generation corridors, areas already zoned for heavy industry, or former coal plant sites with transmission connections. According to Microsoft Copilot, these choices would be the least disruptive to residential areas and could utilize existing infrastructure.
“I’d rather see a data center replace an abandoned steel mill than a farm or subdivision,” the AI said.
It also argued that data center construction should be avoided next to residential developments, schools, parks, and historic districts, unless local voters explicitly supported it. It suggested the Great Lakes region and Midwest as potential regions to build.
Despite the AI’s suggestions, Microsoft’s data center expansion plans have at times faced pushback from local communities concerned about issues including energy use, water consumption, noise, environmental impacts and potential strain on public infrastructure. For example, in Caledonia, Wisconsin, Microsoft decided not to proceed with data center plans after facing significant community pushback.
When reached for comment, Microsoft referred Newsweek to a blog about its AI practices, which included a plan that detailed the company would pay utility rates high enough to cover the data center electricity costs, minimize water use and “replenish more water than we use,” and won’t ask municipalities to reduce property tax rates, among other initiatives.
Google Gemini
Gemini, Google’s artificial intelligence assistant, also argued that new data centers should be built in existing, retired coal plants.
“The current model—concentrating massive “mega-campuses” in a few suburban areas (like Northern Virginia or North Texas)—is creating severe friction over energy prices, noise pollution, land conversion, and water depletion,” Gemini said, then suggested data centers designed for high-density AI model training should be built in Minnesota, Wisconsin, New York, and Michigan, whereas data centers designed for large-scale cloud infrastructure and AI training clusters should be integrated into retired coal plants in Pennsylvania, Ohio, West Virginia, Nevada, Utah, and Idaho.
Newsweek reached out to Google by email for comment.
ChatGPT
OpenAI’s artificial intelligence assistant ChatGPT also outlined environmental concerns and said that new data centers should be built where energy is most abundant, not where it is cheapest. ChatGPT also advocated for building data centers in closed factories or retired power plants to minimize impact.
ChatGPT explicitly said, “keep them away from dense residential neighborhoods.”
Data centers would be best located in the Great Lakes or Upper Midwest, where naturally cooler climates could help assist with cooling, or in the Pacific Northwest.
ChatGPT said it would be more cautious about building data centers in regions where there is severe water scarcity, in dense suburbs, on prime farmland, in flood-prone coastal zones, and areas where the electrical grid is heavily constrained.
Newsweek reached out to OpenAI for comment.
Claude
Claude, a family of artificial intelligence models and a chatbot created by Anthropic, also advocated building data centers “away from residential density,” with the caveat that they should be close enough for workforce housing and jobs to be accessible.
Much of the Southwest should not be considered for new data center builds because it is a water-stressed area, Claude said. Before a data center is built, Claude suggested that independent water-use and grid-impact studies should be completed and published publicly and real penalties should be enforced for noise and light mitigation.
Claude also had something to say about the economic impact of data centers.
“Construction jobs are temporary, and permanent operational jobs per facility are relatively low (often in the dozens to low hundreds for a hyperscale site),” Claude said. “Local governments should negotiate incentives against realistic long-term revenue rather than the inflated projections companies often lead with, and should demand transparency on actual property tax contributions versus abatements.”
Newsweek reached out to Anthropic by email for comment.
What Happens Next?
As artificial intelligence adoption continues to expand, data center development is expected to remain a significant issue for local governments, utilities and residents in many parts of the country.
New proposals will likely continue to face scrutiny over electricity demand, water use, land use and potential impacts on nearby communities, while developers are expected to argue that additional infrastructure is needed to support growing demand for cloud computing and AI services.
How those competing priorities are balanced will vary by community, but questions about where data centers should be built—and what safeguards should accompany them—are likely to remain part of the broader national conversation about the future of AI infrastructure.
Contact Newsweek editors on this story: Samantha Beech and Gray R. Thomas
Related Articles
