The debate will revolve around tariff structures, deviations from zoning regulations, and power transmission corridors. But at its core lies a non-technical political question: Do the people living in this district matter to those who built this building?
The question concerns more than just artificial intelligence. It concerns the very social contract upon which modern democracy is founded.
Democracy has never been a gift from elites who have discovered justice for themselves. Constitutions and elections are the visible mechanisms of limited government. But why did those in power agree to these limitations in the first place? Rights were won by workers, women, soldiers, dissidents, and movements that made dissent politically dangerous, and economic dependence helped make these victories sustainable.
When modern states needed armies made up of ordinary people rather than a warrior caste, they required recruits, taxes, and sacrifices—and thus a certain degree of consent. People who are asked to fight, pay, and suffer for the sake of the state will ultimately ask what place they hold within it.
Similar dynamics were evident in industrial cities, where it became clear that disease knows no class boundaries. Sewer systems, clean water, and quarantine laws reflected the realization that the city’s poor were breathing down the necks of its rich. The elite paid for the well-being of people they did not like because, without them, they themselves could not live well.
Similarly, Henry Ford’s “five-dollar-a-day wage” was not charity. Mass production required mass consumption. Ford recognized a more general truth of the industrial order: wages, consumption, and production volumes were mutually supportive.
In short, since those in power could not prosper without the cooperation of ordinary people, they had reasons to accommodate demands they would have preferred to ignore. AI raises the possibility that this arrangement is weakening.
AI “for the elite”?
It isn’t going away—at least not yet—and it may never disappear in the primitive form envisioned by futurists. This technology may empower more workers than it replaces and create professions that no one has even named yet.
But the political danger does not lie in the fact that every worker will become unnecessary. To undermine the democratic contract, AI does not necessarily have to render ordinary people useless. It need only make the elite less dependent on them.
Old forms of dependence have already weakened. Wars are increasingly fought by professionals, mercenaries, and drones, and decades of automation have led to the emergence of a narrow social stratum considered extremely valuable, a larger group that has lost its stability, and a vast number of citizens who are, in effect, being told that their contribution is not essential.
Suppose that technology fulfills most of the promises made by its proponents, allowing companies to replace—rather than simply supplement—a large number of workers. An automated economy would not lose its consumers as a result; it would be able to sustain demand through exports, government procurement, lending, or transfers to households.
But a society in which most people receive benefits funded by wealth they neither own nor control distributes political influence quite differently than a society in which people earn wages, pay taxes, organize into unions, and negotiate the terms of production.
The latter is rife with conflict, inequality, and exploitation—but it gives citizens bargaining power because they have something to hold onto. A society based on welfare benefits may be richer in material terms, but poorer in terms of social status. It may maintain consumption levels while simultaneously eroding citizenship.
Data centers are not oil wells. But the political economy resembles the “resource curse” long associated with rentier states: when decisive wealth is generated by assets that do not require significant ongoing input from the population, a narrow group of rulers and owners has less incentive to engage in broad negotiations.
The old slogan went: “No taxation without representation.” But the darker reverse is also true: when those who rule and those who own no longer depend on broad-based taxation, labor, or consent, representation begins to look less like a right and more like a cost they would prefer not to bear.
Democratic Control Over Infrastructure
This is precisely why local policy regarding data centers is so important. This industry cannot produce everything it needs on its own. It requires land, water, electricity, and permission from local authorities.
Public resistance is already evident. According to a Gallup poll conducted in March, 71% of Americans opposed the construction of a data center in their region, including 48% who were strongly opposed.
People may view artificial intelligence abstractly with skepticism, but they understand that they face higher electricity bills, threats to their water supply, and an industrial facility that promises only a small number of permanent jobs for the local population.
Permits from local authorities are more than just a hurdle. They represent one of the last forms of influence available to citizens in an economy built around assets they don’t own and systems they can’t scrutinize. This is precisely why the tenuous alliance between Silicon Valley and the populist right is fracturing at the level of power substations, rather than on talk shows.
But local approval is a lever of influence only as long as it must be sought. A government sufficiently committed to the industry’s interests has other tools at its disposal—the power of eminent domain, federal supremacy over local authority regarding facility siting, and emergency powers over the power grid — and U.S. history does not inspire optimism about what lies ahead for communities that own land sought by those in power but do not provide anything those in power need.
The solution lies neither in halting technological development nor in idealizing the industrial order that these technologies may replace. That deal was never fair enough. But if labor ceases to be the primary basis on which ordinary citizens claim a share of national wealth, citizenship will need a different foundation—starting with treating the wealth created by automation as a public issue rather than a private achievement.
This could mean society laying claim to the benefits of automation through taxation and democratic control over the infrastructure on which artificial intelligence depends.
The math here is simple. Every means of pressure ordinary citizens have ever had—the ability to refuse to work, to pay taxes, to send their sons to war, or to vote—presupposes the existence of someone who needs what you can withhold. Where there is no need, there is no leverage. Rights that no longer have a price do not disappear; they become favors granted at the discretion of those who bestow them.
For now, the residents of Northern Virginia still possess one thing that industry cannot create with code: their consent. The question is whether they will use this leverage merely to say “no,” or to demand new terms. Car manufacturers may need fewer workers. Will they be allowed to rely on fewer citizens?
Stephen Holmes,
a professor of law at New York University School of Law and a Richard Holbrooke Fellow at the American Academy in Berlin, is the co-author (with Ivan Krastev) of *The Light That Went Out: The Reckoning, Penguin Books, 2019).
©: Project Syndicate, 2026.
