Tip Sheets
Water Cost Accountability Act is ‘sound’ first step, but leaves much unaddressed
September 3, 2026
Media Contact
The House Energy and Commerce Committee is holding a hearing on the text of the Water Cost Accountability Act this morning, as data centers become a focal point of the campaign trail and public concern about footing their utility bills mounts.
Fengqi You, professor of energy systems engineering at Cornell University, is an expert on data centers’ water and energy footprint and has performed research quantifying the indirect water costs of AI computing.
Fengqi You
Roxanne E. and Michael J. Zak Professor in Energy Systems Engineering
“The Water Cost Accountability Act begins with a sound principle: households and small businesses should not pay the incremental water-system costs created by a large private data-center load. Requiring the data-center owner or operator to cover connection and capacity-expansion costs would be a meaningful ratepayer protection.
“At the same time, the discussion draft is a focused first step rather than a complete water-sustainability framework. It addresses certain public-water infrastructure costs, but it does not by itself require lower water use, evaluate local watershed stress, cover wastewater or electricity-related indirect water use, or establish ongoing public site-level reporting.
“Microsoft’s reported support is understandable because the proposal is consistent with the public commitment Microsoft has made to fund water and wastewater infrastructure required by its facilities. A federal rule could make that cost-responsibility principle more consistent across the industry. The key implementation questions are how incremental costs are defined, independently reviewed, secured against project cancellation or downsizing, and enforced.
“Data-center water use has both direct and indirect dimensions. Water may be consumed on-site for cooling, while electricity generation can create a substantial off-site water footprint. Our study published in Nature Sustainability found that indirect water accounted for approximately 71% of the modeled U.S. AI-server water footprint, so water and energy planning should be considered together.”
