By Steven Cohen, Ph.D., Director of the M.S. in Sustainability Management program, School of Professional Studies
The quietly building consensus against data centers and Artificial Intelligence (AI) needs to be understood and addressed rather than ignored or bulldozed. The President’s unqualified endorsement of data center construction, like his strategy of looting Venezuela’s oil while ignoring its impact on the regime’s legitimacy and greenhouse gas emissions, is predictable and not helpful. I believe that, like the internet itself, AI is inevitable and transformative but must be built with regulations and guardrails. Trying to stop it is futile and destructive. Pausing its development is logical if the pause is used to carefully analyze and mitigate its negative impact. AI should be a tool to enhance productivity and is already serving that purpose, but it could also remove decision-making and the ability to learn from young people, and that must be prevented. It could enhance or destroy education. We must learn how to use it, but must continue to exercise analytic rigor, human values, and judgment. Massive increases in data centers and the energy needed to power them are needed for the current versions of AI, but the need for energy will likely not increase as predicted, since it is a monetary cost that AI companies will want to reduce.
The financial cost to consumers of increased energy use can be reduced by charging data centers more for energy than that charged to nearby residents. The full impact of these centers should be analyzed and mitigated prior to construction. It begins with a complete analysis of the costs and benefits of locating a center. Regulators and developers must conduct an analysis of water use, energy use, and emissions. Developers should be encouraged to train and hire local workers. Government must aggressively and transparently conduct enforcement against companies that break commitments and community benefit agreements. Rural siting may be appropriate in some locations; enforceable community benefit agreements must be part of the deal. Data centers should not be located on prime farmland and should only be built where water supplies are plentiful enough so that the water needed for cooling computers is easily available. Let’s remember that data centers use far less water than agriculture. Developers often prefer rural and exurban locations because land is less expensive, and they typically find there are fewer rules and regulations governing construction than those required by urban sites.
Nevertheless, many older American cities have brownfields, abandoned factories, and underutilized land with industrial-scale infrastructure that could be suitable locations for data centers. They also have access to skilled labor, and their construction would not alter the fundamental nature of a community. Large data centers on rural sites could disrupt small towns and overwhelm local housing, roads, schools, and other resources. An industrial site near the waterfront like Brooklyn’s Industry City could be paired with an offshore wind farm and rooftop solar to reduce fossil fuel use and greenhouse gas impacts from that use. A data center in a city with over eight million people would barely be noticed beyond its neighborhood.
There are some who consider the United States to be in some kind of AI race with China. That is jingoistic nonsense. The goal of AI dominance is idiotic. It is true that nations compete. But it is also true that they create networks of interdependence. It is very difficult for nations to control the science and technology they may invent and first develop. The electric vehicles that China now builds better and cheaper than those built here are being studied and copied by auto companies here and all over the world. Even proprietary AI technology can be studied and imitated. We do not need to respond to the boom in AI and data center investment by abandoning sound governance. Quite the contrary. The public cannot and will not be steamrolled into building data centers. We are already seeing the consensus among Republicans and Democrats leading to a Not in My Back Yard (NIMBY) response to companies that are trying to locate centers by brute force.
In addition to reducing environmental impacts, two of the other pillars of sustainability management are: 1. Workplace and governance equity and fairness and 2. Minimizing negative community impacts. Well-managed organizations adhere to the precept of “mindful management”: paying attention to the organization’s treatment of employees, reducing negative environmental impacts, and maintaining good relations with host communities. Before siting a data center, local stakeholders should be consulted, and the concerns that they raise should be addressed and not ignored. Secrecy and deceptive arguments and behavior by organizations seeking to locate their operations are a mistake. The use of power and economic and political force engenders reactions that can stop development that could have been successfully sited.
Recall Amazon’s failure in 2019 to locate their HQ2 in Long Island City:
Amazon and New York’s shameless governor and New York City’s inept mayor were willing to spend $3 billion to attract one of the richest companies in the world. Local community groups and elected leaders were not consulted and were able to kill the deal. Long Island City is trying to attract development and, if locals had been consulted, Amazon might have succeeded. Instead, they ended up in Crystal City, Virginia and never completed “HQ2.” They did locate 2,000 of their staff in the former Lord & Taylor building in Manhattan, and thousands of others work at other locations in New York City, demonstrating that the $3 billion subsidy was not needed to attract Amazon. New York City’s global and highly educated population attracts businesses; we don’t need to bribe companies. In some cases, companies do succeed in muscling into reluctant communities, and in my view that is an example of poorly thought-through “macho management.” In Amazon’s case, they pulled out of Long Island City unilaterally once they saw the degree of opposition that the proposed subsidy engendered. The premise of their HQ2 siting strategy was to have state and local governments bid against each other to provide siting subsidies. It did not seem to occur to Amazon that this was not an appropriate use of taxpayer dollars anywhere. In the modern era of polarized politics and social media-dominated communications, companies must tread carefully when engaging with communities and stakeholders. Image matters, as one of the world’s richest companies learned in Long Island City.
This is the danger I see with the frenzied effort to site data centers and the huge amount of capital being generated to rapidly develop the computing capacity demand generated by Artificial Intelligence. The damage parents see from the impact of social media on their children was well underway before regulatory guardrails were finally developed. It is unclear if these phone-less school days and return to blue books will benefit children or education. But something needed to be tried. This is fresh in everyone’s mind, and then along comes AI, which appears to be social media on steroids. The corporate push for AI data centers threatens to turn everyone into NIMBY-fueled luddites.
Opposition to data center siting is growing in the United States. An Annenberg survey in August reported that:
“As protesters rally against data center construction in dozens of states and New York and Texas pause the issuance of permits for large data centers, a new nationally representative survey finds that Americans’ opposition to data centers in their communities has risen sharply since last spring. The survey, conducted by the Annenberg Public Policy Center (APPC) of the University of Pennsylvania, finds that about 3 in 5 U.S. adults (61%) oppose the construction of new data centers in their area, up 12 percentage points from a survey fielded in February and March. Americans’ broader views of artificial intelligence (AI), however, have remained stable. The share saying that AI’s impact on the United States over the next 10 years will be somewhat or very negative was 39%, statistically unchanged from the spring. Two-thirds (68%) continue to say the government has done “too little” to regulate AI, unchanged from the spring. The movement over these four months was concentrated in one place: what Americans think about the physical infrastructure of a data center being built near their home.” (Emphasis added)
The shift in public opinion against data center siting is mirrored by state and local proposals and actions to ban or regulate siting. An early September piece in the Washington Post by Chris Hacker, Cole Reynolds, and Klara Auerbach reported that:
“As public backlash erupts against data centers, AI companies have found themselves in a race with state lawmakers. Companies are rushing to construct facilities for the servers that power their platforms. Legislators, across the country and the political spectrum, are increasingly racing to limit them. State lawmakers have already introduced at least 375 bills on data centers this year, according to an analysis of legislative records by The Washington Post. Just over halfway through 2026, that number exceeds the total number of such bills in the three previous years combined. Roughly three-quarters of those bills introduced this year seek to restrict data centers in some way. As with most legislation, a majority of the bills failed to pass, The Post’s analysis found, though many are still pending.”
The image of a race against regulation is accurate and is the heart of the problem. These community fears of data centers and Artificial Intelligence cannot be brushed aside or dismissed. Our federal system of government provides multiple centers of power and gives communities many tools to stop development they oppose. This nation’s housing shortage should be a cautionary tale. NIMBY is a powerful political force that is built on America’s core value of protection of private property. Data center developers should not want to generate opposition from communities who oppose development that locals believe threatens their property and way of life.
The long-term benefit of Artificial Intelligence, like the internet and even social media, requires thoughtfulness and care if it is to be realized. Governments and businesses that ignore the growing opposition to siting data centers do so at their peril. My view is that with care and mindful engagement, data centers can be sited, and AI will continue to be developed. Without such care, the opposite is likely.
Views and opinions expressed here are those of the authors, and do not necessarily reflect the official position of Columbia School of Professional Studies or Columbia University.
About the Program
The Columbia University M.S. in Sustainability Management program offered by the School of Professional Studies in partnership with the Climate School provides students cutting-edge policy and management tools they can use to help public and private organizations and governments address environmental impacts and risks, pollution control, and remediation to achieve sustainability. The program is customized for working professionals and is offered as both a full- and part-time course of study.
