Every year brings a defining moment in technology. For the UK, 2026 is the year its digital infrastructure evolves from a network of data centres into the future engine room of an AI-driven economy. 

For decades, data centres have powered the digital economy. Today, the UK’s data centre structural conversion efforts are so significant that they will change how Britain computes, communicates and competes for years to come. 

However, as enterprise AI adoption becomes the norm, so does the demand for capacity, efficiency and sustainability. To meet this demand, the UK needs a government-industry collaboration to coordinate planning reform, grid upgrades and sustainability standards. 

As Matthew Baynes, Vice President of Secure Power & Data Centres for the UK and Ireland at Schneider Electric, explains in this exclusive interview with Data Centre Magazine: “The question is no longer whether growth will occur, but how intelligently and sustainably we can deliver it.”

Data centres becoming AI factories

The UK is entering its fastest phase of digital infrastructure growth in its history. Plans for nearly 100 new data centre sites are in the pipeline, which speaks to the UK’s desire to move beyond incremental expansion when it comes to its high-density, AI-first facilities.

Around 477 facilities currently operate in the UK, ranking it among the top three global data centre markets. Today, sites across London, the South East, Wales, Scotland and Greater Manchester are being designed as AI factories with large-scale campuses for next-generation workloads, sovereign compute and sustainable energy integration.

“We’re making location, power and cooling choices with national economic implications,” Matthew explains. “The industry needs government fast-track planning and review of national energy pricing strategies for Critical National Infrastructure (CNI) sites, underpinned by joint skills programmes to deploy at scale. That means co-developing national AI infrastructure planning in partnership.”

The UK's AI push has seen the government set up five AI Growth Zones as part of its Modern Industrial Strategy. These zones support large-scale AI data centres by prioritising compute infrastructure alongside clean energy access, skills development and local economic growth. The UK government estimates the programme could unlock up to £100bn (US$135.1bn) in investment while creating thousands of jobs. 

Alongside AIGZs within the AI Opportunities Action Plan, the UK government now deems data centres as CNI, alongside water and gas. “Planning policy, industry strategy and skills must reflect the CNI status”, Matthew says. “Government-industry alignment on grid connection queues and regional development funds will unlock this potential.”

The UK’s data centre design evolution shows the shift from conventional capacity planning to national digital infrastructure planning, where compute and power strategies are inseparable. Site choices, energy sources and cooling architectures now carry strategic economic importance.

“2026 will be defined not just by new builds, but their sophistication if government and industry jointly prioritise power and skills,” Matthew insists.

Ireland’s LEAP: A warning and a blueprint

While the UK accelerates its build-out, Ireland offers lessons in government-industry reset. Known as the home of transatlantic cables and hyperscalers like AWS, Ireland’s data centres power cloud services for millions, supporting 21,000 direct jobs and an ICT ecosystem, employing more than 180,000 people across construction, engineering and specialist services.

However, “that level of success comes with side effects,” Matthew notes. “In Dublin, there’s been grid congestion, long planning disputes and tough questions about energy use, just as AI workloads are ramping up,” he says. 

Europe’s data centre electricity demand could more than double to over 230TWh by 2035 and for Ireland, that projection is significant, because its data centres could account for 30% of national electricity consumption by 2030, the highest share in the EU. 

“Ireland is both a warning and a blueprint,” Matthew says. “It shows what happens when digital growth runs ahead of infrastructure and regulatory clarity but it also shows how to reset the framework.”

This reset is embodied in the Large Energy User Action Plan (LEAP), published in 2026. LEAP is the Irish government’s plan that sets out 17 enabling actions aimed at reopening the market to large energy users (LEUs) – including data centres – under new conditions. 

It promotes a plan-led approach where major investments such as hyperscale facilities, semiconductor fabrication facilities and biopharma plants are steered towards green energy parks – co-located with indigenous renewables like offshore wind and aligned with national priorities for housing, transport and water.

“LEAP shows that government-industry clarity attracts investment while protecting grids. It basically says: we still want strategic digital infrastructure, but it needs to support grid stability, renewables and regional balance,” Matthew explains. “From an investor point of view, that kind of clarity is integral and from a system point of view, it’s essential.”

Why power adequacy is the gatekeeper for AI growth

Behind the UK and Irish stories lies a shared reality: energy infrastructure. “Power adequacy is the gatekeeper for AI,” Matthew explains.

Schneider Electric’s Sustainability Research Institute highlights Adequacy Reserve Margin (ARM) as a key metric for judging whether extra demand can be absorbed without undermining system resilience. Governments set and enforce ARM targets to ensure grid stability.

“Think of ARM as your safety buffer,” says Matthew. “When countries keep that buffer at 15% or more, they can bring on large AI loads with a reasonable degree of confidence. When they sit consistently below it, every new hyperscale or AI campus becomes a risk decision.”

Ireland has experienced this first-hand. Frequent dips below the 15% mark, combined with ongoing reliance on fossil fuels, have made new large-scale connections far more complex to approve and deliver.

“In that context, you can’t just wave through every AI project,” he says. “You need a government-approved framework that asks: does this project help or hurt system adequacy?”

In Ireland, LEAP and the Commission for Regulation of Utilities’ updated LEU Connections Policy do exactly that by tying permissions to system impact. Plan-led siting, co-location with renewables and requirements for flexibility or self-supply all push large users to support rather than strain the grid.

“The UK isn’t in the same position as Ireland, but the direction of travel is similar,” Matthew says. “As AI clusters grow, adequacy-aware thinking will have to play a bigger role. It’s far better to build that into government thinking and market signals now than to retrofit it after the fact.”

How to change burdens into assets

For Matthew, the shift from seeing data centres as potential burdens to viewing them as prospective grid assets is cause for optimism. The EU’s revised Energy Efficiency Directive, which requires operators above 500kW to report KPIs, is already nudging behaviour.

“The days of opaque, one-way consumption are numbered,” he says. “Facilities that flex demand, store energy and return grid value will thrive. Government tax incentives and industry innovation must align here.”

In Ireland, some life sciences operators are already demonstrating what this looks like, with microgrid deployments that combine on-site generation, storage and sophisticated controls to deliver resilience across complex sites. LEAP’s green energy parks are designed to make that type of model scalable: LEU’s acting as prosumers, generating, storing and exporting power in line with system needs.

“Data centres become adequacy remedies,” says Matthew. “UK government-industry partnerships could agree on this model for AI campuses, bringing renewables and flexibility to regions.”

Sustainability at the core, not added on

If power is the gatekeeper, sustainability becomes the condition of long-term legitimacy. 

Ireland illustrates this tension in a snapshot: despite hitting a record 41% renewable electricity in November 2025, the country still faces high energy prices driven by imported gas and oil dependence. Across Europe, electrification rates remain stuck at 21%, with €584bn (US$688.4bn) needed in grids by 2030 to handle growing renewable variability. 

“In an AI context, sustainability is not a line item you bolt on later,” Matthew explains. “It has to shape where you build, how you build and how you operate from the very beginning.”

For him, that translates into AI-first campuses that couple efficient power distribution with advanced cooling and greater focus on heat reuse. It also implies a deeper role for digital technologies and network management tools to cut waste and enable demand-side flexibility.

“Electrification and digitalisation together have the potential to save Europe billions of euros a year in avoided system costs,” he notes. “But that only happens if governments back smarter infrastructure and if operators are willing to run their facilities as part of a more dynamic, data-driven grid.”

“AI leadership will be determined by how well we align planning rules with grid strategy and how we enable data centres to be an active asset to the grid.”

“If industry and government can successfully collaborate we can unlock the AI opportunity delivering job creation and economic upside benefiting both the UK and Ireland.”