The United States will need more than 230 GW of new generating capacity over the next five years, but regulated utilities are expected to add only about 93 GW of accredited supply, leaving a gap of more than 100 GW, according to a new report from Bank of America analysts. The shortfall is driven primarily by AI-powered data centers, which are projected to reshape the nation's electricity landscape after a decade of flat consumption. The report warns that equipment shortages and slow infrastructure buildouts will force data center developers to rely increasingly on on-site gas generation while utilities delay coal plant retirements and scramble to preserve reliability.
Data centers alone could add roughly 125 GW of U.S. electric load from 2026 through 2030, pushing overall electricity demand growth to a 4.1% compound annual growth rate over the period, the report projects. More than 7.5 GW of data center projects with on-site generation are already under construction, with another 60 GW-plus in pre-construction. Large gas turbines, the preferred technology for flexible power, remain largely sold out through 2030 and can take years to enter service after shipment. Coal plants across Maryland, Wisconsin, Indiana, Utah, Kansas, Nebraska and Mississippi have had retirement dates delayed or canceled to preserve dispatchable capacity. A 10% increase in real electricity prices typically results in just a 1% to 2% decline in consumption, according to academic research cited in the report.
The report finds that planned generation additions may overstate available supply because intermittent resources such as wind and solar contribute less accredited capacity during peak demand than their nameplate ratings suggest. BofA analysts note that utilities have revised demand forecasts upward in each of the past three years as AI-related electricity demand has materialized faster than expected. "The market is no longer constrained by demand — it is constrained by where power can actually be delivered," BofA analysts said. Rather than operating entirely off grid, data center facilities with self-generation are expected to combine on-site power with traditional grid connections to improve reliability and shorten project timelines.
The report's forecast comes largely from BofA's semiconductor analysts, whose projections for rapid deployment of AI computing infrastructure underpin the firm's outlook for a widening U.S. electricity supply gap. AI infrastructure is reshaping long-term electricity demand after a decade of largely flat consumption growth driven by efficiency improvements, LED adoption and distributed solar generation. With manufacturing capacity for large gas turbines largely committed through 2030, data center developers are increasingly turning to natural gas reciprocating engines, which can be deployed more quickly and respond rapidly to changing loads. Manufacturers including Caterpillar, INNIO, Rolls-Royce and Wärtsilä have expanded production to meet rising demand, the analysts said.
Battery storage, transmission expansion and regulatory changes that increase utilization of existing generating assets could help address reliability challenges, though the analysts caution that transmission projects often take years to permit and build. They cite the Champlain Hudson Power Express, which took 16 years from planning to energization, as an example of the development timelines facing new infrastructure. The concentration of AI-driven data center growth in the U.S. is prompting utilities and regulators to address how to connect large new loads and allocate the cost of new infrastructure. Higher power prices could drive some demand destruction among customers and energy-intensive industries, but electricity demand is relatively inelastic in the short to medium term, the report concludes.