The UK has committed £1.5 billion to its next wave of AI infrastructure, but the question now is whether this investment will be spent effectively.
The traditional approach to meeting rising compute demand has been to build new facilities from scratch, a model that is proving outdated. This approach involves breaking ground on a greenfield site, waiting years for planning and grid connection applications to be approved, and hoping the facility is ready before the workloads it was designed for become obsolete.
However, a faster and cheaper way to build AI infrastructure has been demonstrated by Samir Tabar, chief executive of AI infrastructure company WhiteFiber. Tabar's company has focused on acquiring underutilised industrial sites that already have the necessary power capacity and proximity to major metro areas.
WhiteFiber's flagship US project, a former textile mill in Madison, North Carolina, was acquired for a fraction of the cost of comparable greenfield land and shell development. The site was converted into a hyperscaler-grade AI campus in just over a year, and a ten-year, roughly $865 million colocation agreement with European AI hyperscaler Nscale was secured.
The economics of this deal are significant, as it compressed years off the delivery timeline, sidestepped cost overruns and planning delays, and provided a high-quality solution. In an industry where speed is a decisive competitive variable, this is not a marginal advantage, but the whole game.
The UK, however, is still relying heavily on the greenfield playbook, even as it invests in AI compute through its AI Growth Zones and national supercomputer commitments. Grid connection queues remain a significant bottleneck to getting new capacity online, and there are many underused, well-powered sites that could be brought into service through a retrofit-first strategy.
If the UK wants its £1.5 billion to translate into operational AI capacity rather than years of planning applications, it should study the retrofit model demonstrated by WhiteFiber in North Carolina seriously.
