Data Center Boom Faces Reality Check on Water and Costs

The data center boom runs on water and price.

In a Lawfare discussion, Andy Masley joined editors to challenge the industry’s most repeated claims about the footprint of data centers, from water use to land take and the price of power. Masley insists that growth projections and siting narratives are too often built on misleading or incomplete data, and he pushes for a clearer, evidence based accounting of what these facilities really require and cost. The conversation, guided by Kevin Frazier of the University of Texas AI Law Program and Lawfare’s Alan Rozenshtein, centers on what Masley calls scaling laws for data centers: rough, real world limits that shape where and how big these facilities can, and should, become.

Water is a central battleground. Masley notes that cooling strategies and regional water availability drive wild variation across jurisdictions, with some regions able to spare consumptive use while others face tight supply and higher scrutiny. The panel does not pretend data center water use is negligible, but it argues that the totals are highly context dependent and often misrepresented in lobbying or investor pitches. For compliance teams and operators, the upshot is a push toward credible, location specific metrics rather than blanket assurances. In practice, that means regulators and developers alike should demand transparent water intensity data, cooling technology choices, and local drought risk assessments when evaluating new builds or expansions.

Land use and community impact comprise the other axis of Masley’s critique. Large parcels are a hallmark of data center campuses, but the local effects, such as traffic, noise, and the perceived bargain of tax incentives, generate a divide between public perception and the economics of siting. Masley argues that communities deserve hard numbers on land use footprints and long term cost to residents, not vague promises about jobs or philanthropy. The discussion acknowledges that while data centers can bring grid connected reliability and tax revenue, without careful planning and solid data, towns risk chasing incentives that create more cost than benefit over time.

Electricity rates, a perennial flashpoint in siting debates, also come under examination. The panel stresses that data centers are grid dependent firms with unique demand profiles, and wholesale power pricing plus local rates can shift the financial calculus of a project. The implication for operators is clear: proximity to robust, affordable, and reliable power matters, but so do ancillary costs such as demand charges and capacity auctions. For policymakers, the lesson is to avoid simplistic narratives about cheap power everywhere and instead design rate structures and performance incentives that reflect actual consumption patterns and the broader grid needs.

From a practitioner perspective, several concrete takeaways emerge. First, institutions should adopt standardized, location specific resource metrics that capture water use, land footprint, and energy demand over time, rather than rely on headline claims. Second, siting decisions must weigh not only upfront incentives but long range costs to communities and the grid, including potential future upgrades and the risk of stranded assets. Third, cooling choices matter more than marketing suggests: air cooling versus liquid cooling has different implications for water needs, land use, and maintenance. Finally, regulators and industry alike should demand transparent risk disclosures tied to climate and drought forecasts, so that scaling plans align with the realities of water and power markets.

The session signals a broader move toward credible, data driven debate over how fast data centers can, or should, grow. As Masley puts it, the future lies not in grandiose claims but in the stubborn math of scaling laws, where resource constraints and local conditions ultimately govern the shape of the footprint.