This piece is authored by Oliver Inderwildi, University of Cambridge, and Thomas Le Goff, Assistant Professor of Law & Technology at Télécom Paris – Institut Polytechnique de Paris.
At a time when Europe’s electricity grids are already stretched thin, traditional web searches are being replaced by artificial intelligence (AI) queries that require ten times more energy! And, it doesn’t stop there: Technological change is accelerating – driven by feedback loops we couldn’t have foreseen just a few years ago.
Innovations are disrupting economic sectors and rippling through their value chains – from decentralised finance and cryptocurrencies to streaming platforms and the rise of artificial intelligence. A fair share of these disruptions is digital in nature, making the growing digital economy not just a trend, but a defining factor for future competitiveness.
The digital economy relies on robust digital infrastructure; this infrastructure, in turn, depends on one critical enabler: a stable, secure, and scalable energy supply. Here’s the catch: While digital innovations proliferate at unforeseen rates, the electricity system that supports it is subject to long lead times due to – among other things – significant regulatory burdens and decarbonisation targets. This discrepancy in agility, combined with both sectors’ relevance for competitiveness, makes the seamless integration of the digital and energy infrastructures a highly complex topic.
Data centres (DCs) are a cornerstone of this infrastructure. In 2021, CERRE’s report on “Data Centres & the Grid” explored the complexities of integrating data centres into these intertwined—but unequally paced—economic systems1. It asked a fundamental question: How can the electricity grid support a booming digital economy while staying on course to meet emission reduction targets? A mere three years late and we are navigating a very different landscape.
What’s Changed Since 2021?
In short, a lot has changed: Digitalisation continues to accelerate, energy prices and security are taking centre stage again and everyday AIs (large language models or LLMs) are installed on most smartphones, all while grids are stressed and their development is naturally slow. As an example, traditional web queries are being replaced by LLM queries at an energy expense almost ten-fold higher (Figure 1, orange)! Consequently, the landscape has been turned upside down. Goldman Sachs projects that by 2030, AI alone will be responsible for 20% of electricity demand from data centres globally (Figure 1, blue).
Figure: Projected global electricity demand from data centres (blue), with the growing share of AI workloads shown in light blue – reflecting the rapid rise in AI adoption. One illustrative example of this shift: traditional web searches are increasingly replaced by AI-powered queries, which require up to 10 times more electricity (orange).
Moreover, the European Green Deal’s targets for decarbonisation and digital transformation have to be balanced; digital technologies are a double-edged sword as they are significant electricity utilisers while simultaneously supporting decarbonisation endeavours in all economic sectors2. Moreover, edge computing is on the rise; its decentralised data processing will relieve grid demand while creating more access points making grid management even more complex.
All this increases the complexity of an already delicate balancing act between digitalisation, decarbonisation, and economic competitiveness.
Key Issues of this Evolved Landscape
There are overarching trends that shape the technical demands on the grid while testing the regulatory framework intended to support it, the list below outlines the key factors.
1. System-Level Considerations (Overarching Trends)
- The acceleration of digital innovation, often with mutually reinforcing technologies, increases modelling uncertainty and complicates long-term infrastructure planning.
- Aligning increased electricity demand from digital infrastructure with national and EU-wide decarbonisation targets remains a major challenge.
- Technology trends like edge computing introduce a new layer of complexity by e.g. shifting computation closer to demand reducing data transport but increasing the number of grid nodes.
- The current geopolitical context has revived the debates on energy security and digital sovereignty in Europe amplifying the importance of resilient energy & digital infrastructure.
- Increasing regulatory burden on both digital (AI Act, Cyber Resilience Act, etc.) and energy (EU Energy Efficiency Directive, EU Renewable Energy Directive) developments with calls to simplify in the name of competitiveness.
- Global regulatory trends, esp. U.S. deregulation, is pressuring EU states and challenging legal cohesion3. In response, local approaches to data centres vary—some restrict them due to network strain, others promote them at environmental cost. This fragmentation makes a unified EU policy hard to achieve.
2. Specific Challenges & Opportunities
Next to the overarching trends outlined above, there are specific technical and regulatory challenges when integrating DCs into the electricity grid that could provide ultimately provide opportunities for grid flexibility and resilience.
- Demand shifts & increases: e.g. through the high energy requirements of AI-driven data centers or the 24/7 operation of conventional DC, represent a significant change in demand patterns. These shifts must inform grid design to prevent congestion, local grid stress, or instability in energy supply.
- Regulatory lag: The regulatory framework was not designed for the current demand shifts for energy and DC connections, creating uncertainty and administrative burden. Evolved regulation must support the integration of DCs into the grid while avoiding inflexibility and inertia.
- Infrastructure Development & Integration: Connection rules differ between member states, and consequently connection queues have been formed due to network congestion. The sharing of best practices in terms of connection agreements and regulatory approval must be shared among stakeholders.
- Grid Flexibility (temporal & spatial): DCs could enhance grid stability by providing temporal flexibility e.g. by timing low-latency workloads to use excess baseload availability or providing back-up electricity generation and storage. Moreover, intelligent location of DC – e.g. between low-carbon electricity sources and demand centres – could increase spatial flexibility while reducing network latency.
Both overarching trends and specific challenges can be addressed in such a way that they created a path with opportunities.
The Path Forward
The questions outlined above illustrate that the interplay between digital growth and energy constraints is becoming ever more important. It’s about developing an infrastructure that safeguards Europe’s energy security while achieving climate targets and ensuring economic competitiveness. Regulation has to balance different targets efficiently while supporting this evolution – a feat that most certainly will not be easy.
By addressing the questions outlined herein, an updated and holistic picture of this rapidly evolving landscape will be created and concrete policy recommendations for intelligent integration of DCs into the electricity system will be deducted. This blog marks the start of a six-month project to revisit and expand the 2021 CERRE report on Data Centres & the Grid.
As Europe races to achieve its climate and competitiveness goals under the Green Deal, understanding how data centres interact with the energy system—and how to govern this intersection—has never been more urgent. Watch this space.
1 CERRE, “Data Centres and the Grid – Greening ICT in Europe“, 2021
2 O.R. Inderwildi & M. Kraft, “Intelligent Decarbonisation”, SpringerNature 2022
3 MLex, “Germany’s new government may ask for EU AI Act revision”, March 2025
