Voters rejected AI data centers over resource concerns. Here's why measuring water usage is hard
Voters approved 40 of 41 ballot measures seeking to restrict artificial intelligence data centers because of concerns about water use and electricity demand. The article explains that water consumption varies widely depending on cooling systems, climate and hardware, with examples from Equinix, Meta and SpaceX facilities.
Voters across the United States approved 40 of 41 ballot measures proposing restrictions on artificial intelligence data centers, reflecting growing concerns about their demands on local water supplies and electric grids. CNBC reports that operators often disclose water consumption companywide rather than by facility, making site-level impacts difficult to measure.
Data center researcher Shaolei Ren of the University of California, Riverside, and Equinix sustainability head Christopher Wellise explained that newer AI systems require liquid cooling, although many facilities use closed-loop systems that recycle water internally. External heat-rejection systems can still consume substantial water during hot weather or require significantly more electricity when using dry cooling.
The article cites Meta's planned 1-gigawatt facility in Lebanon, Indiana, Equinix's facility in San Jose, California, and SpaceX's Colossus data centers near Memphis, Tennessee. Meta and SpaceX are also associated with water-treatment or recycling projects, while SpaceX's on-site gas turbines have faced lawsuits alleging Clean Air Act violations and noise pollution.