An article by BCS Consultancy’s Nick Marlow, Head of Project Management.
Executive Summary
- AI is creating huge waves across Europe’s data centre delivery ecosystem, with JLL forecasting AI will represent half of all data centre workloads by 2030.
- BCS Consultancy’s Data Centre Truths 2026 report shows that 78% of respondents have seen a notable uplift in demand linked directly to AI over the past 12 months, highlighting how quickly this shift is materialising.
- AI is fundamentally reshaping data centre operations in Europe, but the industry’s success will be determined by how decisively it responds.
While much of the public narrative continues to frame AI as a software or technology trend, its most immediate impacts are being felt in physical infrastructure. Across Europe, AI is creating tangible delivery challenges. Power availability, cooling capacity and sustainability are no longer secondary optimisation considerations. They are now primary constraints shaping site viability, programme certainty and long-term operation.
AI workloads are exponentially growing with a strong impact already felt across the data centre delivery ecosystem. JLL forecasts that AI will represent half of all data centre workloads by 2030 as both public and private organisations are looking to AI to increase efficiency and productivity. In the UK alone, the government is targeting £45 billion in efficiency savings by automating routine tasks and enhancing public services. More tellingly, BCS Consultancy’s Data Centre Truths 2026 report shows that 78% of respondents have seen a notable uplift in demand linked directly to AI over the past 12 months, highlighting how quickly this shift is materialising.
That momentum is even more pronounced closest to delivery. The same research found that 100% of developers and investors, and 91% of service providers, have experienced an increase in AI-driven demand. This aligns with wider enterprise trends, where confidence in AI output continues to grow. While many deployments remain in pilot or early-stage implementation, larger organisations are now scaling investment as AI tools move into everyday workflows.
This growth trajectory raises fundamental delivery questions for the industry. How many facilities will be required, where can they realistically be delivered, and can the industry supply them at pace? This is translating directly into a significant increase in data centre investment across Europe with the EUDCA projecting €100 billion of investment in the region by 2030. The challenge now is whether operators, designers and policymakers are equipped to meet the infrastructure demands of AI.
Power availability and the regionalisation of AI data centres
Power requirements for AI data centres are one of the key issues impacting data centre growth in Europe. With power applications now reaching over a gigawatt, this scale of demand cannot simply be transported into traditional FLAP-D markets without significant grid upgrades and infrastructure investment.
As a result, Europe is seeing a clear regionalisation of AI data centres away from these historic hubs. Fortunately, there are several power generation projects currently looking for investment and demand while they wait for grid infrastructure to catch up in these markets. This has created opportunities for AI-ready developments in new locations, provided that power solutions can be structured effectively.
One of the most prominent responses to this challenge has been the emergence of “bridge to grid†solutions, utilising green power and gas through microgrid configurations. These approaches are helping to meet short-term demand while longer-term grid reinforcements are delivered. However, government policy around gas power generation needs reviewing if these solutions are to be unlocked at scale and used alongside BESS, solar and wind options.
Cooling, sustainability and delivery risk are now inseparable
As cooling strategies evolve to meet AI demands, sustainability has moved from simply environmental performance metrics into delivery risk considerations. Energy use, water consumption and legacy infrastructure constraints are now directly linked to programme certainty, planning outcomes and long-term operational resilience.
Across much of Europe, water use is a critical planning issue. Cooling solutions that rely heavily on water abstraction face increased scrutiny from regulators and local authorities, particularly in regions that already experience water stress. Similarly, legacy infrastructure that was not designed to support sustained high-density AI workloads introduces long-term operational risks that cannot be mitigated through incremental upgrades alone.
These sustainability challenges affect whether projects can operate as intended over their full lifecycle, and perhaps more importantly, whether or not they receive planning consent at the start. Cooling decisions made early in the design process now have implications that extend far beyond mechanical efficiency, influencing everything from site selection to long-term reputational risk.
Europe’s distinct regulatory and political landscape
Europe faces a distinct set of challenges and opportunities when scaling AI-ready data centres. Unlike the US and China, which are leading the race to supply the world’s stock of AI data centres through agile deregulation and investment incentivisation, Europe operates within a far more complex regulatory environment.
Promises of deregulation across EMEA have often been met with additional bureaucracy and uncertainty around AI growth zones and power gateways. Much of this regulation remains unproven, and its application across different jurisdictions has slowed the pace of supply at a time when demand is accelerating. This uncertainty increases delivery risk and makes it more difficult for developers and investors to commit capital with confidence.
At the same time, European governments are increasingly focused on AI sovereignty. Our own government interactions have demonstrated that governments across EMEA are estimating demand for both public and private use, while also reviewing the stock of data centres within their sovereign borders. In the context of global politics, there is a growing preference for data centres to be located domestically, further reinforcing the need for scalable, deliverable solutions within Europe.
Supply chain pressure and the skills challenge
The rapid expansion of AI-driven infrastructure is placing unprecedented strain on the supply chain. Main contractors are growing quickly, often under significant quality and time pressures. Although the EMEA construction market is entering a period of recovery and growth, it continues to face structural challenges, including a shortage of skilled labour and inflationary pressures.
Insurance markets are also adjusting, as they seek to understand and price the construction values and risk profiles associated with gigawatt-scale campuses. This has introduced further uncertainty into project costing at a time when capital discipline is already under pressure.
Inflationary pressures in the US, particularly around wages and materials linked to the rollout of large AI campuses, are expected to ripple into EMEA markets. As projects that were previously put on hold begin to ramp up alongside new developments exceeding 200MW, competition for resources will intensify.
The skills gap will only increase as demand accelerates. Delivering on the promise of AI will require a huge influx of people across multiple sectors, not just engineering and construction. Accountants, investment bankers, lawyers, manufacturers, automation experts, mining specialists, architects, contractors, strategic advisors, HR professionals, coaches and mentors will all be needed. The list goes on.
A delivery-led mindset is now essential
AI is fundamentally reshaping data centre design and operations in Europe, but the industry’s success will be determined by how decisively it responds. Cooling and sustainability are no longer optional design considerations; they are delivery-critical constraints that will dictate whether AI infrastructure can be built at scale. The decisions taken now will determine whether Europe can deliver AI infrastructure at scale, or whether ambition continues to outpace physical reality.
