Chasing Power, Not Proximity: Europe's AI Data Centres Migrate to the Hinterlands
New JLL data shows hyperscale campuses scheduled for 2026 to 2028 will sit an average of 175 kilometres from major cities, nearly four times farther out than projects delivered in the previous three years, as developers chase available energy and affordable land over urban access.
LONDON: Europe's artificial intelligence data centre pipeline is moving away from the continent's traditional hubs, with developers increasingly willing to trade proximity to users for cheaper power, faster grid connections, and larger parcels of land. The shift is sharpest among AI training campuses, which need enormous amounts of electricity and water for cooling but do not need to sit close to the populations they serve.
JLL, one of the world's largest commercial real estate firms, shared data with Reuters showing that the average distance from a major hub for hyperscale facilities delivered between 2022 and 2025 was 46 kilometres. For projects scheduled to come online between 2026 and 2028, that figure jumps to 175 kilometres. Greenfield projects now account for 39% of Europe's future pipeline, compared with just 8% of completed builds, while the share of inner-city projects is expected to fall from 13% to 5%.
"The determining factor is increasingly where sufficient power can be secured, rather than simply where demand exists," said Assad Noori, JLL's head of data centres in Europe, the Middle East and Africa. "Data centres are being brought to where the power is, not the other way around."
The data from DC Byte supports that view. Of nine proposed gigawatt-plus data centres across Europe, only one is planned near a major city, Paris. The rest are scattered across areas ranging from rural Spain to northern Sweden. Core markets such as Frankfurt, London, Amsterdam, Paris, and Dublin remain the largest and continue to draw enterprise demand, but land shortages, planning restrictions, and lengthy grid queues are pushing the hyperscale buildout outward.
The cost differential is substantial. Powered land in core European markets averages €2.36 million per megawatt of IT load, while secondary cities such as Copenhagen, Warsaw, and Milan come in at €978,000. In tertiary areas such as Bordeaux, the average drops to €512,000 and can go as low as €200,000. Amsterdam remains the most expensive market at about €2.7 million per megawatt, followed by London at €2.6 million and Frankfurt at €2.5 million.
JLL estimates the world's four largest hyperscale cloud providers will spend $725 billion in 2026, up 77% from $410 billion in 2025, with most of that going to AI computing and data centre infrastructure. By 2030, AI workloads could account for roughly half of global data centre capacity. The migration to cheaper land and energy can direct investment into underprivileged regions that need jobs, but it also creates friction. Residents in those areas are increasingly worried about natural habitats, water competition, and the strain large facilities place on local power networks.
BuiltWorld AI Operational Take: The move to greenfield sites at 175 kilometres and beyond changes the risk profile for project delivery in Europe. Remote locations solve the land cost and grid queue problems, but they introduce longer transmission line buildouts, thinner construction labour markets, and water access issues that may not be fully priced into early site selection models. Teams that can secure power agreements in secondary and tertiary markets before competitors arrive will lock in a structural cost advantage, but they will also need to budget for community engagement and ecological mitigation that core urban projects never had to face.