AI's Power Hunger Drives Up Costs for Households

A report by Lynn Parramore at the Institute for New Economic Thinking details how the massive energy demands of AI data centers are being passed on to consumers through higher electricity bills, rent, and borrowing costs. Hyperscale data centers, costing around $11 billion each, are proliferating, with Virginia hosting roughly 300. Utilities often bundle grid upgrade costs into general rates, as seen in the PJM power grid where regular customers covered $4.3–$4.4 billion in transmission upgrades. In Virginia, data centers account for nearly 40% of electricity use, potentially raising household bills by $14–$37 monthly. Texas, Georgia, Arizona, and Ohio are emerging as new hotspots. Economist Servaas Storm warns that higher energy prices fuel inflation, leading to higher interest rates. Critics like Darren Bush advocate that cities should sometimes reject data center projects due to externalized costs. Ilya Sutskever, former OpenAI chief scientist, predicted in the 2019 documentary iHuman that the Earth's surface could be covered with solar panels and data centers.