For not having any legs, datacenters are leaving a huge footprint. Tech’s AI boom is driving a massive US data center buildout, but a new study warns that the environmental costs could be immense if expansion isn’t carefully managed. As companies like Meta and OpenAI commit trillions to infrastructure, energy and water demands are bound to skyrocket, particularly in states with already stressed grids or scarce water resources. The analysis highlights states like Texas, Montana, Nebraska, and South Dakota as optimal sites, balancing renewable energy availability with sufficient water for cooling. However, traditional hubs like Virginia and California face challenges meeting sustainability goals amid continued growth. Researchers caution that net-zero pledges may be unrealistic as AI demand outpaces efficiency gains, potentially adding tens of millions of tons of carbon dioxide annually. The planet doesn’t need more hot takes, just careful planning to scale AI infrastructure safely.