# Gelsinger Warns U.S. Power Limits Could Slow AI Expansion By Simon Yoon Canonical URL: https://www.tokenpost.com/news/technology/29595 Published: 2026-10-10T16:10:13.000Z Updated: 2026-10-10T16:10:13.000Z Section: Technology > The former Intel CEO said energy capacity has stayed nearly flat for 10 to 15 years as data centers compete for electricity. Pat Gelsinger warned that limited U.S. power availability could slow artificial-intelligence and data-center expansion as electricity demand rises alongside other infrastructure constraints. The former Intel CEO and Playground Global general partner said U.S. energy capacity had remained effectively flat for 10 to 15 years as coal generation was retired and renewable capacity was added. He made the comments in a podcast published Oct. 9 at 6 a.m. ET (10 a.m. UTC). Gelsinger estimated that annual energy-capacity growth had recently reached about 4%, but said the pace was insufficient for expected AI and data-center demand. He also said gas-turbine delivery times had reached approximately eight years and predicted more data-center project defaults if electricity supplies are not secured. U.S. electricity consumption had been nearly flat for 15 years before rising at an average annual rate of 2.1% over the previous five years. Consumption is projected to grow 0.9% to 1.6% annually through 2050, with data-center server use a major driver. Electricity sales are forecast at 4,135 billion kilowatt-hours in 2026 and 4,211 billion kilowatt-hours in 2027. The power constraint is emerging alongside other limits involving GPUs, memory and advanced chip packaging. AI facilities require large, concentrated electricity supplies, while new projects also depend on generation capacity, transmission access and grid-connection timelines. The same constraints affect cryptocurrency-mining facilities, which can be large electricity users and face similar power, pricing and permitting challenges. [Earlier coverage examined the U.S. data-center power gap](), including potential additions from gas turbines, fuel cells, nuclear power and converted crypto-mining sites. The rising electricity needs of AI, data centers, manufacturing and crypto mining are adding to the demand for expanded generation and transmission capacity. ## Links in this article - [Earlier coverage examined the U.S. data-center power gap](https://www.tokenpost.com/news/technology/27314)