Thought Leadership
Why Europe’s multibillion-euro AI gigafactories could take years longer to materialize despite ample funding and sustained demand.
The European Union plans to invest €20 billion in building five artificial intelligence (AI) gigafactories. Public debate has focused on demand, economic viability, and locations, while overlooking an uncomfortable reality within Germany’s power grid: securing a firm grid connection can take seven to ten years. Even if planning began immediately, the first gigafactory could not realistically come online before 2033. The real weakness in Europe’s digital strategy is not demand. It is the capacity to execute.
We are currently seeing the same pattern across several active projects. It is no longer an isolated problem: A data center project is fully financed and approved, supplier contracts have been signed, and construction equipment is ready to move in. Yet the project still cannot proceed. The reasons are almost always the same:
- Grid operators cannot meet binding commitments for power delivery.
- Design changes, such as changes to cooling technology, come too late, while increasingly long lead times for major equipment put overall project schedules at risk.
- Established general contractors are fully booked, while traditional local construction companies often reach their limits because they lack the specialized expertise these projects require.
This is the reality behind a debate that, from Berlin to Brussels, is currently focused on the wrong questions. Industry voices, including those cited by Handelsblatt on May 10, have warned of overcapacity and questioned the economic viability of the five projects backed by €20 billion in funding. The criticism is that Europe is building without secured demand.
The market data, however, paints a different picture. According to the latest edition of BCS Consultancy’s annual Data Centre Truths report, which surveyed more than 3,000 industry participants across 41 countries, 93 percent of respondents expect demand to continue rising. The real problem lies in an area that has received little attention in the public debate so far.
Well Funded, Yet Still Failing
Capital and political announcements may be the price of admission to today’s market, but neither guarantees that a project will actually be delivered. Well-funded and highly publicized developments regularly stall because several rarely discussed factors converge: grid connection capacity arrives late, design changes are made too far into the process, equipment lead times are long, and critical expertise and skilled personnel are lacking across the entire project ecosystem.
A major development in the Frankfurt area, for example, recently failed not because of insufficient capital or technical limitations, but because of a lack of dialogue with a local citizens’ initiative that ultimately brought enough political pressure to stop the project. The common denominator behind these delays is neither money nor technology. It is a lack of expertise and experience across all relevant stakeholders at the same time, from developers and planners to utilities, authorities, and construction companies.
Capital Is Growing Faster Than Execution Capacity
Two figures illustrate just how pronounced the imbalance between available capital and operational delivery capacity has become. On one side, 95 percent of surveyed experts expect the supply of qualified professionals to shrink further. Half of all developers already report having to postpone milestones in ongoing projects as a result.
On the other side is an investment boom. According to the European Data Centre Association’s State of European Data Centres 2026 report, investment is expected to rise from €7.7 billion in 2024 to €25.7 billion by 2028, more than tripling within just a few years. The capacity of suppliers, grid operators, and, above all, qualified professionals is nowhere near keeping pace. The core problem can therefore be reduced to a simple formula: Capital and ambition are growing much faster than our ability to build and operate the infrastructure.
The Infrastructure Will Not Be Ready Until the 2030s
The real limiting factor for the planned gigafactories lies elsewhere: grid connections.
Current waiting times in Germany for a firm commitment to connect a major project to the power grid range from seven to ten years. By comparison, under optimal conditions, completing a gigafactory and bringing its first phase into operation takes roughly three years from the start of planning. That means the long wait for a grid connection could push a gigafactory planned today into operation in 2033 or even later, unless it is built on land where power capacity was secured speculatively years in advance.
Based on the current situation, gigafactories may therefore not be able to contribute meaningfully to European AI sovereignty until the middle of the 2030s, rather than by 2027, as politicians frequently promise. The only real alternative would be a fundamental reform of how grid connection capacity is allocated, combined with a major acceleration of the entire process.
Why Networks Could Be the Answer
This reality has an important strategic implication that has received little attention in the public debate: Gigafactories should not be designed as isolated projects. Instead, they should form part of a broader network linking large centralized facilities with smaller, locally based data centers. Such a network architecture has a critical operational advantage. Smaller, locally focused data centers can be built more quickly and can often be developed in urban areas on sites that already have basic infrastructure and grid connections in place.
A European strategy built around a network of large and small facilities could bring initial capacity to market sooner than one relying exclusively on five centralized gigafactories. Greater connectivity and decentralization would also improve resilience against disruptions across the wider system and distribute the benefits more evenly across Europe instead of concentrating them in just five locations.
What Needs to Happen Now
For this network-based approach to work, the industry needs an honest assessment of its current position, along with two major policy reforms. The first should establish at least nationwide standardized approval procedures for data centers in Germany. Local political processes and, in some cases, overstretched municipal planning authorities are currently causing significant project delays. The second concerns the expansion of transmission and distribution grids, as well as a shift in capacity allocation away from a “first come, first served” model toward a “first ready, first served” approach.
A “first ready, first served” model would allow grid operators to focus on projects that are genuinely capable of moving forward rather than tying up capacity and administrative resources with speculative requests. Such a reform would not require billions in additional spending, yet it could fundamentally and sustainably change the speed at which data center projects are completed across Europe. The industry also needs to take a critical look at itself, particularly when it comes to the skills shortage.
Too often, qualified professionals are simply recruited away from competitors instead of companies investing in training and long-term skills development. BCS Consultancy has taken a different approach through its own established training program. Across the wider market, however, far more new talent needs to be developed instead of repeatedly competing for the same limited pool of experienced professionals.
Those Who Can Deliver Will Win
Europe does not have a demand problem, nor does it lack ambition. It has an execution problem.
That cannot be solved with another multibillion-euro funding package. It requires a sober reassessment of priorities. Anyone serious about advancing Europe’s digital infrastructure should focus less on isolated flagship projects and more on a broadly distributed network, replace political announcements with reliable commitments in grid connection registers, and move beyond the cannibalization of skilled personnel in favor of structured workforce development.
The decisive players in the years ahead will not be those controlling the most capital. They will be the ones capable of delivering locally, predictably, and at industrial scale. That capability cannot simply be decreed. Europe must build it systematically.
