The Data Center Debate Is Missing the Point

The fastest-growing asset class in real estate is running into its first real constraint. Not capital. Not demand. Trust.

As Maine enacts the nation’s first statewide moratorium on large data centers and communities across the country push back on new projects, the industry is hitting a wall built on eroded trust. In Maine, lawmakers did not just debate a project. They stopped an entire category of development.

Demand for computational power and storage (aka compute)  is accelerating at a pace the physical world is struggling to match. Artificial intelligence, cloud infrastructure and digital services are driving one of the largest infrastructure buildouts in modern history.

Yet, projects are being delayed, contested and increasingly canceled as local opposition intensifies.

The conversation has fractured into two opposing narratives. On one side, data centers are framed as economic engines that generate tax revenue and attract investment. On the other, they are portrayed as extractive infrastructure that consumes land, water and energy while offering limited local benefit.

Both sides are partially right. And both are missing the point.

The real issue is not whether data centers should be built. It is how they are being built, and who is included in that process.

The industry is answering the wrong question.

One of the biggest mistakes in the current debate is the assumption that all data centers are the same. They are not. A hyperscale artificial intelligence campus in a rural, water-stressed region operates very differently from an urban colocation facility or a latency-sensitive edge node. Their energy intensity, cooling strategies, workforce needs and community impacts vary significantly. Hyperscale AI campuses can require massive power and water inputs with relatively limited permanent jobs, while urban colocation and edge facilities tend to operate with smaller footprints and closer alignment to local demand.

When those differences are ignored, the industry ends up defending itself against the worst version of itself.

This lack of differentiation has given critics an advantage. The most extreme examples shape public perception, while the industry responds with generalized defenses.

At the same time, the industry continues to rely on arguments that do not resonate locally. National statistics about GDP and job creation are frequently cited as proof of value. They are impressive. They are also insufficient.

Residents are not debating GDP. They are asking something much simpler.

What happens to my electricity bill? Where does the water come from? How much noise will this create? How many permanent jobs will exist after construction ends? What does this project give back to the community, and when?

A trillion-dollar industry argument does not answer a household’s utility bill.

If those questions are not answered clearly and in enforceable terms, the macroeconomic case will not carry the project.

That is the mistake. The industry did not lose ground on technology or capital. It lost ground on governance. The path forward is not persuasion. It is governance.

Without enforceable frameworks that define how costs, resources and benefits are shared, opposition will continue to grow. Communities are asking for specific commitments tied to measurable outcomes.

If the industry wants to regain legitimacy, it needs to start answering the right question.

It means large energy users operating under dedicated rate structures that prevent cost shifting. It means minimum take agreements that ensure infrastructure investments are paid for. It means binding community benefit agreements that specify local hiring and investment. It means clear disclosure of water sourcing and usage. It means enforceable noise standards and limits on backup generation.

These are not theoretical ideas. They are already emerging. In Northern Virginia, data centers have generated outsized tax revenues relative to local service demand. In Texas, recent agreements have included water usage caps and reporting requirements. Community benefit agreements in multiple states are beginning to formalize local hiring and infrastructure contributions.

At the same time, the scale of impact is becoming harder to ignore. Data centers already account for roughly 4 percent of U.S. electricity consumption, with projections pointing significantly higher.

Against that backdrop, the industry must confront an uncomfortable reality. Its approach to transparency has become a liability. The use of nondisclosure agreements, redacted resource data, and accelerated approvals has not protected projects. It has undermined trust. Secrecy is not protecting projects. It is creating opposition.

The rise of organized opposition reflects that shift. What was once fragmented has evolved into a coordinated, bipartisan and increasingly effective movement. Projects are being delayed or blocked at scale.

This is no longer a fringe reaction. It is a structural force shaping where and how data centers can be built.

Ignoring it is no longer an option. Adapting to it is now a requirement. The next phase of data center development will not be determined by demand. It will be determined by whether the industry earns the right to build.

The winners in this next phase will not be those who move fastest. They will be the ones who answer the right questions.

And trust, unlike compute, cannot be scaled overnight.

Suhail Y Tayeb is clinical assistant professor at New York University’s Schack Institute of Real Estate and director of the Center for the Sustainable Built Environment.