President Donald Trump said his administration will drive AI corporations to shoulder the energy costs of large-scale data facilities, claiming that households have to not face higher utility bills due to fast infrastructure buildouts.
“I never want Americans to pay higher Electricity bills due to Data Centers,” Trump wrote on Truth Social, adding that big tech companies “ought to ‘pay their own way.’”
Microsoft Is First to declare New Commitments
In a Truth Social post, Trump stated that Microsoft could define concrete steps “later this week,” framing it as a model for how major AI infrastructure players need to take care of power costs. Microsoft’s declaration landed on January 13, 2026, with the corporation committing to pay better utility rates supposed to fully cover the cost of its electricity use, while also working with local utilities to amplify supply.
It also pledged to release data center water-use and replenishment data by U.S. Region and replenish more water than it consumes. The move comes as local and state-level backlash grows round data center expansion, specially in area already hosting dense clusters of facilities.
Why Power Costs Have Become a Flashpoint
Data centers already represent a meaningful slice of U.S. electricity demand. A U.S. Department of Energy report found data facilities consumed approximately 4.4% of total U.S. electricity in 2023 and projected they might reach roughly 6.7% to 12% by 2028 (with a wide usage range relying on buildout pace and efficiency gains).
Forecasts range, but multiple estimates location U.S. data center electricity use in the 300–400 TWh/1year range by 2030. For context, total U.S. electricity consumption in 2024 became about 3,953 TWh.
Globally, the energy pressure is also growing. The International Energy Agency venture electricity demand from data centres international is set to more than double by 2030 to around 945 TWh, with AI a major drive.
What to Watch Next
Trump’s feedback and Microsoft’s plan point to a broader policy query: how utilities and regulators should allocate the prices of new generation, transmission, and distribution upgrades required to serve AI-era load growth—without shifting those costs onto residential customers.
