Trump links AI data centers to utility rate protection
The AI Power Struggle: How States Are Fighting for Control Over the Grid
The race for artificial intelligence dominance is fueling an unprecedented demand for electrical power, and as data centers become the engine of this revolution, the fight over who pays the price—and how that power is managed—has reached the state level. In a move aimed at tackling spiraling electricity costs, President Donald Trump’s “ratepayer protection pledge” was recently signed by 23 state governors and 187 utility companies and data center developers. This agreement is Washington’s attempt to rein in energy prices driven by the massive power appetite of AI infrastructure.
This promise emerged after the President previously ordered major AI hyperscalers, including Meta, Amazon, and OpenAI, to commit to generating their own power. The core objective was to bring down electricity costs, with officials promising more affordable and reliable energy for American families. However, the immediate impact of these high-level commitments on utility bills across the nation has been minimal.
Despite the federal pledge, the cost crisis remains acutely felt at the local level. PJM Interconnection, which oversees the nation’s largest power grid, faced a significant challenge when it slapped Maryland with a $2 billion bill to upgrade its grid just to accommodate out-of-state AI data centers. Independent watchdog groups have pointed out that the 75.5 percent increase in power costs seen in the U.S.’s largest power region is directly attributable to these booming data center demands.
This disconnect between federal promises and regional reality highlights a fundamental tension: while the goal is affordability, the massive infrastructure buildout seems to be outpacing regulatory action. The administration is pushing for acceleration of AI adoption and infrastructure development, seeking to maintain global supremacy in the technology race, but this push-forward is colliding with growing public opposition.
The backlash is real. Concerns regarding electricity costs, water quality, noise pollution, and environmental impact caused by data center expansion are causing widespread public friction. In response, local jurisdictions are increasingly stepping into the fray, using their regulatory powers to slow or halt development.
These efforts have manifested as rapid, temporary bans on new data center developments across the country. Jurisdictions like Seattle, which hosts giants like Amazon and Microsoft, are moving swiftly to implement moratoriums to study the community impact of these massive buildouts. Similarly, the state of New York has enacted a one-year ban on data center projects larger than 50 megawatts, signaling a shift in power away from unchecked development toward public scrutiny.
Further demonstrating local action is Oregon, which passed the POWER Act in April 2025—nearly a year before the federal pledge was announced. This law forces large electricity consumers to reflect the true costs of their service. As a result, Portland General Electric (PGE) increased data center power bills by 30 percent while simultaneously cutting residential electricity costs by 1.3 percent, ensuring that development carries its full financial weight.
The ultimate success of the ratepayer protection pledge hinges on whether these commitments can be legally enforced and whether developers will accept responsibility for their infrastructure costs. While the administration insists that states must support data center growth, the reality on the ground is a decentralized tug-of-war between massive technological ambition and local community demands for responsible resource management. The future of AI infrastructure depends not just on powerful chips, but on successfully balancing unprecedented demand with public accountability.