President Donald Trump used his State of the Union address to confront a growing political vulnerability: rising electricity prices tied to the rapid buildout of AI infrastructure. As he framed it, his proposed solution places the cost burden on large tech companies building energy-intensive data centers.
Under what he called a “ratepayer protection pledge,” Trump said major tech firms would be required to supply their own electricity for new AI facilities rather than drawing additional power from already strained regional grids. The proposal, light on details, is clearly designed to reassure households facing higher monthly utility bills.
Federal data shows residential rates increasing by roughly 8 percent in 2025, intensifying voter frustration over rising living expenses ahead of the midterm elections. While the administration has emphasized expanded oil and gas production, consumer power bills have continued to trend upward.
Data centers are, to be sure, a significant contributor to that pressure. Somewhere between 650 to 700 facilities are in development nationwide. Collectively, they would require power comparable to hundreds of large generating units. Some data centers devour as much electricity as mid-sized cities.
Trump’s plan would require companies to construct dedicated generation assets, such as natural gas plants, to serve their own operations. Energy Secretary Chris Wright said the administration has been in discussions with leading AI developers and described the talks as productive. He added that companies may also contribute funding to grid upgrades.
Several tech firms have already signaled support for measures that limit cost impacts on residential customers. Microsoft and Google have previously announced commitments to cover the infrastructure expenses associated with their data center buildout.
A Complex Issue
The economics of AI infrastructure are more complex than Trump’s pledge suggests. Even if data center operators build standalone generation, they often remain connected to the grid for backup and reliability. That connection can trigger new investments in transmission lines and distribution systems, costs that are typically shared across ratepayers.
Grid operators across the US have approved billions of dollars in new high-voltage projects, with data centers among the largest beneficiaries. Those expenses are spread across tens of millions of customers.
Additionally, utilities may upgrade substations and reinforce networks in anticipation of data center demand that ultimately fails to materialize. In such cases, regulators must determine who absorbs the stranded costs.
Critics of the administration argue that policy shifts have worsened the affordability problem. Since returning to office, Trump has scaled back federal support for wind and solar programs, ended electric vehicle tax incentives, and relaxed environmental regulations affecting power plants. Environmental groups contend that limiting renewable expansion could restrict supply growth at a time when demand is accelerating.
Balancing Global Competition and Affordability Issues
Beyond economics, the energy issue influences global AI competition. The White House is a major cheerleader for AI development, viewing AI leadership as crucial to compete with China. Officials insist that cost-sharing arrangements with industry will allow the US to expand computing capacity without eroding public support.
Trump is trying to strike a balance, Joseph Hoefer, Principal and Chief AI Officer at Monument Advocacy, told Techstrong.it. “He wants to accelerate the buildout of AI infrastructure so the US stays ahead in the global race, but they also understand that voters are sensitive to rising electricity bills. It’s not a pivot in his AI strategy, but an effort to scale it responsibly while keeping the focus on American competitiveness.”
