As data center developers compete for a cut of the hundreds of billions of dollars flowing into the artificial intelligence industry, the queue to join the UK’s power grid has become jammed with projects that will likely never get built. This snarl is exacerbating already years-long wait times for viable projects and messing with attempts to forecast energy demand and plan grid expansions.
The Ofgem Proposal
In July, the UK’s energy regulator, Ofgem, laid out a proposal meant to force "phantom" data centers out of the swollen queue. Under the plans, developers would be required to put down a steep, nonrefundable deposit that could balloon to hundreds of millions of dollars for the largest projects. Developers would also be required to line up customers in advance and prove they have the funding to complete their builds.
Ofgem faces a "Goldilocks" conundrum: reforms must be burdensome enough to deter speculators but not so much that they drive legitimate data center projects to other shores, undermining the UK’s ambitions to meet the voracious demand for compute required for AI.
From 41 to 125 Gigawatts
The queue began to balloon toward the end of 2024, around the time the government designated data centers "critical national infrastructure." Between November 2024 and June 2025, the total energy demand of projects in the connection queue increased from 41 gigawatts to 125 gigawatts. New data centers make up 73 gigawatts of that—equivalent to one and a half times the peak demand for the entire UK last year.
The crowded queue is largely the product of an incentives problem. Previously, it cost only a few thousand dollars to join the queue, leading to "no-downside bets" by developers. This has also attracted middlemen who apply for power with the aim of later flipping land at a premium, a practice described as "scalping for concert tickets."
Risks to Competitiveness
Industry experts believe Ofgem’s reforms risk making the UK—already unattractive for high energy costs and land scarcity—one of the world’s most expensive places to build a data center. There is also concern that the steep fees could stifle growth by forcing out smaller, AI-specific operators that may be unwilling or unable to provide such large sums at the development stage.
