Data Drain: The Land and Water Impacts of the AI Boom (www.lincolninst.edu)

The AI boom is driving a rapid expansion of hyperscale data centers that voraciously consume land, electricity and freshwater. Large facilities can use up to 5 million gallons of water per day and draw as much electricity as 100,000 homes; in 2023 U.S. data centers used about 176 TWh (roughly Ireland’s national consumption) and that could double or triple by 2028. Regionally concentrated clusters—Northern Virginia alone hosts ~300 sites—and projects like Meta’s Hyperion (expected to exceed the power draw of New Orleans) illustrate how siting decisions, tax incentives and corporate investments are fueling growth. Projections in Texas estimate data center water use rising from 49 billion gallons in 2025 to as much as 399 billion gallons in 2030. Chip fabrication also uses huge volumes of ultrapure water, compounding supply pressure. Those resource demands have concrete local consequences: paved campuses displace farmland and habitat, transmission lines and gas plants increase emissions, diesel backup generators create air-quality problems, and evaporative cooling withdraws water that largely never returns to local watersheds. Economic benefits are often temporary (construction jobs) even while municipalities grow dependent on volatile tax revenue. Technical mitigations exist—closed-loop cooling, immersion cooling, siting in wetter/cooler basins, and water replenishment programs—but they raise costs or electricity trade-offs and are not yet widespread. The story highlights an urgent need for transparent reporting, local planning and policy to balance AI infrastructure growth with ecological and public-health constraints.