One monthly bill Americans can’t avoid is quietly surging thanks to emerging industry: data
Americans are likely to see higher electric bills in the years ahead as the fast-growing artificial intelligence industry reshapes how much electricity the country needs, according to new research from the Federal Reserve Bank of Dallas.
The Dallas Fed estimates that data centers already operating across the United States have pushed wholesale electricity prices up by 2% to 6% on average, with even sharper increases in regions where these facilities are clustered together.
A single large data center can consume as much power as a small city, according to figures compiled by Fed researchers. The electricity is used to run the vast networks of computers that power AI applications, which are being adopted by businesses and consumers at a rapid pace.
Researchers at the Dallas Fed modeled a middle-range scenario in which the cost of generating electricity could climb 20% to 30% higher by 2028 than it would have without new data centers coming online. The figure refers to generation costs rather than retail rates, and the researchers caution that the impact on household bills will depend on a range of factors, including how utilities and regulators allocate the costs of new infrastructure.
Wholesale electricity makes up only part of what consumers ultimately pay. Energy generation accounts for roughly half of a typical retail electricity price, with the remainder covering transmission, distribution and other expenses. Wholesale increases also tend to reach household bills gradually rather than all at once.
The strain on the grid is beginning to influence policy at both the state and federal levels.
President Donald Trump has pushed to expand America's AI infrastructure while also backing a voluntary pledge aimed at preventing data centers from driving up household utility bills. The dual approach reflects the tension between encouraging the growth of a strategic industry and protecting consumers from rising costs.
In Texas, Republican Gov. Greg Abbott has ordered regulators to pause data center projects seeking to connect to the state's main power grid until they undergo a comprehensive audit. Pennsylvania Democratic Gov. Josh Shapiro has also moved to tighten oversight of large-scale data center development as his state weighs how to balance new investment with rising demand for electricity.
New York Democratic Gov. Kathy Hochul has gone further, imposing a one-year moratorium on new hyperscale data centers. Other states are weighing similar steps as concerns grow over electricity prices, grid reliability and the speed at which new facilities are being built.
The political attention comes as affordability, energy policy and the future of AI infrastructure have become central issues heading into upcoming elections. Lawmakers and regulators across the country are grappling with how to accommodate the enormous energy needs of AI while keeping electricity affordable for households.
As more data centers connect to the grid, utilities may need to build additional power plants, transmission lines, substations and other infrastructure to serve them. Who ultimately pays for those upgrades depends in part on how regulators and utilities divide the costs between the data centers themselves and other customers.
The Dallas Fed's research highlights a basic but consequential reality: AI requires enormous amounts of electricity, and the country's power system is already feeling the effects.
