Oct. 15, 2026
In recent months, the data center discussion has evolved into a maelstrom of opposition, with Gallup reporting that a large majority of Americans oppose siting a data center in their area; 48 percent say they are strongly opposed.
This puts planners and local decision-makers in the difficult position of finding a way to reconcile vociferous and emotional local concerns with behind-the-scenes legal requirements.
4 principles to know
Demands to restrict, ban, or even incentivize data centers need to fit within a long-established legal framework of equitable utility provision, pricing, and local land-use regulation. While this varies by state and region, there are several principles planners should know.
1. THERE ARE LEGAL LIMITS TO BANS
All states require utilities to supply power and water to all prospective customers, which prevents discrimination against individuals or classes of use. Whether proactively or reactively, these laws have prohibited denial of services, for example, to communities of color and immigrants. Protection is typically couched in language invoking the provision of safe, adequate, efficient, and reasonable service. Demands that local governments “just say no” to data centers by withholding water and power supply hook-ups violate this foundation.
Relatedly, efforts to alter the local development code by categorically banning data centers may breach state property rights protections. A 2025 Texas law largely prohibited moratoria on specific land uses, and a 2026 one-year moratorium adopted by the state’s rural Hill County was rescinded after a developer filed a $100 million lawsuit.
2. IT'S OK TO LET INFRASTRUCTURE CATCH UP
Utility connections can only proceed as quickly as infrastructure allows, so a data center developer can’t speed up the process by filing a lawsuit if there is no infrastructure to support that high-load use.
Cities and counties may legally delay permitting while studying and adapting to the new power and water demands posed by these users, such as building power transmission and distribution infrastructure, extending recycled water pipelines, or developing additional water sources. That process may also occur at a larger, grid-level analysis, such as for an entire state, like in Texas.
Still, evidence must support slowing the permitting process to allow infrastructure to catch up to growth, and it must provide a path to accommodate new users. Even when these large-load users commit to developing their own on-site power or water source, they must still comply with state and local regulations.

The Physical Footprint of Artificial Intelligence explores the physical effects of AI deployment and highlights core considerations for zoning updates and areas of opportunity for land-use decision-making processes.
3. RETHINK UTILITY RATES
In application, these considerations are complex. For electricity, industrial users (which include data centers) are typically assigned lower rates due to their efficiency. Because they use a lot of high-voltage power, they generate a high revenue return for the utility compared to residential and commercial customers. But the explosive proliferation of data centers disrupts the rate equilibrium. While their rising demand may drive up the price of power and water, the most significant cost impact is infrastructure — like new delivery lines and pipes.
Several states, including Virginia and Oregon, are now proposing that large-load users be identified as a discrete class in these studies, allowing for greater nuance in determining the price of their impact on the system. This path must follow legally mandated processes rather than simply assuming higher service costs, which would be challenged as illegal and inequitable. Relatedly, the concerns that data center demand is inflated, leading to sunk costs for infrastructure construction that is never needed, can justify policies requiring potential new users to cover their own infrastructure costs in advance, as New Jersey has recently done.
4. JUSTIFY ZONING DECISIONS
Cities that choose to better regulate the location and features of data centers may be asked to recommend specific design guidelines or changes to planning and zoning codes. As many municipalities discovered too late, their local codes did not define or contemplate data centers, leaving no foundation for enforcing restrictions and allowing them to appear in seemingly inappropriate areas.
But new rules developed to remedy this should be carefully crafted to avoid legal challenges. While concerns are reasonable, they may not align with legally settled justifications for land-use segregation. The key is a reasonable connection between their purported impact and any restrictions imposed. For example, relegating data centers solely to heavy industrial zones may not make sense if a city allows land uses that are demonstrably more noxious (through noise and chemical use) in light industrial zones. The city must back up that argument with data, or the large-load user can claim inequitable treatment.
Design regulations should be consistent with the existing regulatory context. If the code prohibits windowless buildings from districts meant to promote mixed use and walkability, for example, then it is reasonable to apply that restriction to ban featureless data centers in those areas. Conversely, adding that restriction to a district solely because a data center wants to locate there would be suspect.
While these limitations can prove frustrating — and may thwart citizen demands for quick, effective policy actions — their purpose is sound: preventing discrimination against arguably unpopular users and equitably apportioning the costs of growth.

