CU Boulder report outlines lawmaking strategies to regulate AI data centers
A CU Boulder report outlines strategies states are pursuing to regulate AI data centers' water use, estimating significant increases in consumption. While state-level legislation in Colorado has failed for three years, local municipalities like Longmont, Boulder County, and Broomfield have enacted bans or moratoriums on new data centers due to environmental concerns.
A new report from the University of Colorado Boulder, titled "Thirst for Data: State Responses to Data Center Water Use," details four main strategies states are using to regulate the water consumption of AI-driven data centers: reporting, conservation, incentives, and prior appropriation requirements. Co-authored by Daniel Anderson, a water law fellow at the CU Law School, the report highlights growing concerns among local communities and lawmakers regarding data centers' massive energy and water use, potential water contamination, pollution, noise, and land use.
The report estimates that direct water use by U.S. data centers more than tripled from 5.6 billion gallons in 2014 to 17.4 billion gallons in 2023, with projections showing it could nearly double again to 32.8 billion gallons by 2028. Indirect water consumption for power generation was approximately 211 billion gallons in 2023. While Colorado's state legislature has failed to pass AI data center legislation for three consecutive years, local municipalities are taking action.
Longmont City Council has banned hyperscale data centers within city limits, implementing new restrictions to prevent large energy-intensive facilities. Boulder County enacted a six-month moratorium in June on new data center applications to develop code changes, potentially prohibiting them in industrial areas. Similarly, the Broomfield City Council unanimously passed an 18-month moratorium to consider new regulations for the city's code. Anderson noted that this is a rapidly evolving policy area, with few states having explicitly passed policies regarding data center water use.