Data center developers are increasingly building their own power plants directly on their properties to bypass the years-long wait times required to connect to the public electricity grid. This shift allows them to keep pace with the rapid construction of new facilities for artificial intelligence, which often outstrips the ability of traditional utility companies to expand their infrastructure, according to the authors of the newly released report.
“In some areas, AI infrastructure is expanding faster than electricity networks can connect new load. Lengthy grid connection queues and grid expansion timelines are colliding with data centre delivery cycles of one to two years, turning electricity availability into a binding constraint on deployment.”
Essentially, it takes a much longer time to build new power lines and substations than it does to set up a new data center. Because companies are racing to launch new AI services, they cannot afford to wait five to ten years for a standard grid connection. By installing their own equipment—often small gas turbines or engines—they can generate the power they need immediately and start operations years earlier than if they relied on local utilities.
The report “Key Questions on Energy and AI” was published in April 2026 by the International Energy Agency in Paris, France. Part of the World Energy Outlook Special Report series, the analysis was prepared by a team led by Thomas Spencer and Siddharth Singh under the direction of Laura Cozzi. The publication provides a comprehensive assessment of the rapidly evolving intersection between artificial intelligence, data center power demands, and global energy markets.
