AI data centers strain local grids and water, sparking regional opposition

The conflict centers on who bears the costs. AI data centers consume far more power than traditional ones, drawing electricity around the clock and generating heat. Bloom Energy’s January 2026 report forecasts U.S. data center IT load doubling from about 80 gigawatts in 2025 to 150 gigawatts by 2028. PJM’s grid operator saw capacity auction prices surge from $28.92 per megawatt-day in 2024/25 to $329.17 in 2025/26, with 63% of the increase attributed to data center demand. Virginia residential prices rose about 13% over 12 months, and nearly three-quarters of Virginia voters blamed data centers for the spike. Meanwhile, water use remains a local concern: Arizona faces a 760,000 acre-feet cut in Colorado River use, prompting Tucson to reject Amazon’s Project Blue and Chandler to block another proposal. Google’s planned campus in Botetourt County, Virginia, could use up to 2 million gallons of water daily, straining regional supplies.

Industry advocates argue data centers bring economic benefits, such as $1.3 billion in property taxes paid by Northern Virginia data centers in 2024 and $2 billion in local revenue. However, job creation is limited—Brookings found data centers add only 100 to 200 jobs over a decade, with no wage impact. States like Texas and Maryland offer tax breaks tied to job creation thresholds, but critics argue these incentives often come with nondisclosure agreements that exclude public oversight. Meanwhile, companies like xAI (now owned by SpaceX) have built temporary gas turbines to bypass grid delays, sparking lawsuits over air pollution. The broader debate hinges on whether costs—like grid upgrades and water infrastructure—should be shared equitably or shifted to data center operators.