AI Data Centers Move Away From European Cities for Power, Land

European AI infrastructure shifts to the periphery

Europe's artificial intelligence data center developers are abandoning traditional urban locations in favor of rural and secondary markets where power is abundant, land is affordable, and grid connections move faster, according to commercial real estate data analyzed by Reuters.

Between 2026 and 2028, new hyperscale data centers will sit an average of 175 kilometers from major European hubs—more than three times the 46-kilometer average distance of facilities built between 2022 and 2025. The shift reflects a fundamental constraint: cities like London, Frankfurt, and Amsterdam are running out of both suitable land and available electrical capacity to support AI training campuses that can require gigawatts of power.

Why it matters

This geographic redistribution could reshape regional economies across Europe, channeling billions in infrastructure investment to areas that have historically struggled to attract technology capital. But it also sets up potential conflicts with local communities concerned about environmental impact, water consumption for cooling systems, and strain on regional power grids. The trend signals that AI's infrastructure demands are outpacing the capacity of established technology hubs, forcing a fundamental rethinking of where digital infrastructure gets built.

Greenfield projects dominate the pipeline

Greenfield developments—built from scratch on undeveloped land—now account for 39 percent of Europe's future data center pipeline, compared to just 8 percent of recently completed projects, according to data from JLL, one of the world's largest commercial real estate firms. Meanwhile, inner-city locations have dropped from 13 percent of delivered projects to only 5 percent of planned facilities.

Of nine proposed gigawatt-scale data centers across Europe tracked by DC Byte, only one is planned near a major city (Paris), with the remainder spread across locations from rural Spain to northern Sweden.

Economics drive the exodus

The cost differential is stark. Powered land in core markets averages €2.36 million per megawatt of IT load, compared to €978,000 in secondary cities like Copenhagen and Warsaw, and just €512,000 in tertiary markets such as Bordeaux—where costs can drop as low as €200,000 per megawatt. Amsterdam leads European markets at roughly €2.7 million per megawatt, followed by London at €2.6 million and Frankfurt at €2.5 million.

"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 centers for Europe, the Middle East, and Africa. "Data centers are being brought to where the power is, not the other way around."

AI workloads reshape capacity planning

JLL estimates the world's four largest hyperscale cloud providers will spend $725 billion in 2026, up 77 percent from $410 billion in 2025, with most capital directed toward AI computing and data center infrastructure. By 2030, AI workloads could consume roughly half of global data center capacity.

Martin Jensen, president of JLL's EMEA data centers division, noted that while Frankfurt, London, Amsterdam, Paris, and Dublin remain critical for enterprise demand, "hyperscale AI infrastructure requires a completely different scale of power and land."

The details were first reported by Reuters correspondents Simon Jessop and Iain Withers.