What happened

TechCrunch has reported that the rapid expansion of data centres, driven by heavy investment in AI infrastructure, is running into real-world obstacles that could prevent many announced projects from ever being completed.

The obstacles

The publication identifies three specific hurdles: widespread local opposition to data centre construction, permitting delays, and a scramble for electricity supply. These are practical barriers that can stall or kill projects regardless of how much funding has been promised.

The deeper problem

Beyond those obstacles, TechCrunch notes a more fundamental issue. Many data centre projects that have been publicly announced may never get built at all, even as billions of dollars in AI infrastructure funding is committed. The gap between announcement and actual construction appears significant, though the reporting does not provide specific figures on what proportion of projects are at risk.

What remains unclear

The available reporting does not specify which projects are most at risk, how many announced developments have already been cancelled, or what proportion of committed AI infrastructure funding is tied to projects that may not materialise. Without that data, the scale of the gap between announcements and reality cannot be assessed.