Electricity costs rise amid data center boom

Electricity costs are rising nationwide, with a 6.5% average retail residential price increase between May 2024 and 2025, driven significantly by the proliferation of power-hungry data centers for AI. This surge is leading to higher consumer bills, with grid operators like PJM passing billions in infrastructure costs (an estimated $9.3 billion) to customers, and some states experiencing over 30% increases. Utilities face challenges as data center demand may outstrip supply, while also risking over-investment in new transmission if the AI boom's energy needs are overestimated, potentially impacting future growth and utility stability.

Analysis

A surge in electricity demand, primarily driven by the proliferation of AI-powering data centers, is directly translating into higher energy costs for consumers and creating new capital risks for utility providers. U.S. retail residential electricity prices saw a nationwide average increase of 6.5% between May 2024 and May 2025, with more acute spikes in specific states like Maine (+36.3%). The cost pressure is not solely from immediate consumption but also from forward-looking infrastructure investments by grid operators like PJM, which covers Mid-Atlantic and Midwest states. An analysis by the IEEFA attributes a significant portion of a recent PJM capacity market price increase, valued at $9.3 billion, to anticipated data center load growth, with these costs being passed on to the broader customer base. This dynamic is particularly concentrated in regions like Northern Virginia's "data center alley," where a legislative commission forecasts potential monthly bill increases of $14-$37 by 2040. A critical risk is emerging for utilities: they are undertaking substantial capital expenditure to build new transmission capacity for the AI boom, which could result in stranded assets and financial distress if the projected demand fails to materialize, turning the boom into a bubble.

Key Decisions for Investors

• Investors in the utilities sector should exercise caution with companies heavily exposed to regions with rapid data center growth, scrutinizing capital expenditure plans and the risk of stranded assets should AI-driven energy demand fall short of forecasts.
• Consider opportunities in companies that provide energy efficiency solutions for data centers, as the theme of rising electricity costs creates a strong tailwind for technologies that can mitigate power consumption and operational expenses.
• Monitor regional utility performance and regulatory filings, as the disparity in rate increases (e.g., Maine's +36.3% vs. Nevada's -17.7%) highlights that impacts are not uniform and depend heavily on local grid capacity and data center concentration.
• Pay close attention to capacity market auction results from grid operators like PJM, as these serve as a leading indicator for future rate hikes and the financial costs being underwritten to support the AI infrastructure build-out.