Artificial intelligence companies have risen quickly to become a defining trend in global markets. Even relatively young European AI firms have seen valuations multiply within months. NVIDIA’s ascent to the world’s most valuable company illustrates the economic potential emerging across the AI value chain. Yet each step of this value chain depends on one core foundation: energy. While public debate centres on models, training costs and possible market overvaluation, the structural bottleneck is shifting from computing power to energy supply.
Stay informed – subscribe to our newsletters
This shift became clearer in late 2025, when Google acquired energy startup Intersect for around USD 4.75 billion. The deal reflects a broader trend: technology companies are no longer just securing chips and data centres, but also energy infrastructure.
From data centres to “AI factories”
Data centres are evolving into industrial production facilities. NVIDIA now refers to them as “AI factories”: energy goes in, tokens come out. Training large models consumes hundreds of megawatt-hours of electricity, and inference workloads are rising rapidly. As AI adoption accelerates, data centres are creating new load profiles for energy systems: highly concentrated, near-continuous baseload at a limited number of grid connection points.
Transatlantic AI collab signals new phase in EU digital energy
In some regions, data centres already account for double-digit shares of electricity consumption. Grid connection capacity and “time-to-power” are becoming strategic competitive factors for AI infrastructure. Cost structures are also shifting. While accelerator chips remain the largest upfront capital expenditure, electricity is increasingly dominating operating costs. Estimates suggest that 40–60 percent of operating costs for AI-intensive data centres are related to electricity.
AI meets the fully electrified economy
In some regions, data centres already account for double-digit shares of electricity consumption. Grid connection capacity and “time-to-power” are becoming strategic competitive factors for AI infrastructure. Cost structures are also shifting. While accelerator chips remain the largest upfront capital expenditure, electricity is increasingly dominating operating costs. Estimates suggest that 40–60 percent of operating costs for AI-intensive data centres are related to electricity.
Artificial intelligence: The key to transforming renewable energy systems?
AI is accelerating this trend by digitising and automating production and control systems, and in many cases digitalisation means electrification. The additional demand from AI acts as a multiplier on already rising electricity consumption. Efficiency improvements in chips and software will reduce the energy required per computation, but this is likely to trigger a rebound effect: lower costs enable new applications, which in turn drive up total demand. Absolute electricity consumption is therefore likely to continue to increase.
Energy systems as the new scaling layer
For investors and energy market participants, this represents a structural shift. Value is moving from purely digital AI models towards solutions at the intersection of energy markets, grid infrastructure and digital control systems. Energy has always been a strategic asset, but the nature of value creation is changing. In an AI-driven, electrified economy, value is determined not only by installed generation capacity, but increasingly by the ability to manage and balance the system intelligently.
While AI increases electricity demand, it can also help operate a more complex, renewable-based energy system. As variable renewable generation expands across Europe, energy systems require more sophisticated forecasting, optimisation and control. AI-based software can coordinate distributed energy assets, improve demand-side flexibility and optimise energy consumption across industry, buildings and transport. Companies such as Encentive, DeltaCharge and FLEXeCHARGE are applying AI to optimise industrial demand, EV charging and asset aggregation, helping monetise flexibility and stabilise decentralised power systems.
Lessons from chess – strategic thinking for solar leaders
A growing number of European energy-tech companies are addressing these challenges. Encentive, for example, uses AI to optimise industrial electricity consumption based on price and renewable forecasts, enabling companies to monetise flexibility. In the mobility sector, DeltaCharge applies predictive charging algorithms to commercial EV fleets, reducing peak loads and energy costs. Platforms such as FLEXeCHARGE aggregate decentralised assets like batteries and EV chargers into virtual power plants that help stabilise increasingly decentralised electricity systems.
Europe in the competition for AI infrastructure
Europe enjoys several structural advantages as a location for AI infrastructure and data centres. The continent combines a strong industrial base, world-class research institutions and rapidly expanding renewable energy capacity. Major data centre clusters have emerged across the region – Frankfurt, Amsterdam, Paris, London and Dublin – often referred to collectively as the “FLAP-D” markets. These hubs benefit from dense fibre connectivity, proximity to major internet exchange points and access to large electricity markets. Frankfurt, for example, hosts DE-CIX, one of the world’s largest internet exchange points. At the same time, Europe is developing ambitious AI ecosystems, including initiatives such as the Innovation Park Artificial Intelligence.
Qubite taps desert sun to power European data infrastructure
However, Europe faces structural challenges in scaling energy-intensive AI infrastructure. Data centre developers often contend with long grid connection timelines, limited transmission capacity in high-demand regions, complex permitting procedures and comparatively high electricity prices. In several European markets, grid access has become a critical constraint for new data centre projects. As a result, “time-to-power” is emerging as a decisive factor in location decisions, alongside connectivity and access to talent.
This dynamic places energy infrastructure at the centre of Europe’s competitiveness in the AI economy. Expanding renewable generation, accelerating grid upgrades and enabling flexible demand management will be vital to accommodate AI-driven electricity demand and the broader electrification of industry, mobility and heating. For investors and infrastructure developers, the message is clear: AI investment increasingly follows energy availability.
Energy the foundation of the AI economy
The key insight is simple: AI will not only be limited by computing power, but increasingly by available and controllable energy. Energy is not competing with AI, but enabling it. For investors and the energy sector, this creates a long-term opportunity at the intersection of digitalisation and decarbonisation. The real value will lie not only in AI models, but in the intelligent orchestration of a fully electrified economy. (Felix Krause/hcn)
