AI data centers could push household electricity costs higher, study finds

A new analysis from the Federal Reserve Bank of Dallas links the rapid expansion of AI data centers to higher electricity prices for U.S. households. The study finds that data facilities added in recent years have already raised average wholesale power rates by roughly 2% to 6% across the country, with especially sharp increases in regions hosting many of these sites. Under a middle-range projection, the cost of generating electricity could be 20% to 30% higher by 2028 compared with a scenario without the new centers, though retail bills will not jump by that exact margin because wholesale costs represent only about half of the price consumers pay.

The report notes that utilities may need new power plants, transmission lines and substations to serve the growing demand, and the allocation of those upgrade costs will depend on regulatory decisions. Politicians such as Donald Trump, Greg Abbott, Josh Shapiro and Kathy Hochul have taken steps ranging from voluntary pledges to moratoriums and audits to curb the impact on consumers. The issue is becoming a focal point in debates over energy affordability, grid reliability and the pace of AI-related investment.