AI Data Centers Must Support Clean Energy Goals
The European Commission is signaling that Europe’s AI infrastructure strategy will be tied closely to the EU’s energy, climate, and environmental objectives. Energy Commissioner Dan Jørgensen stated that companies seeking to benefit from the AI boom in Europe must show that their data centers support the bloc’s transition to clean power, including renewables and nuclear energy, rather than increasing reliance on fossil fuels.
This position comes as Brussels prepares a new tech sovereignty package aimed at reducing dependence on foreign providers in cloud computing, AI, and semiconductors. Data centers are expected to become a major stress point for Europe’s electricity system, with energy consumption from the sector potentially doubling or more over the next decade. For EU digital policy, this links AI competitiveness directly with energy regulation, infrastructure planning, and climate compliance.
The Commission is also working on a sustainability label for data centers, expected to assess facilities on issues such as energy efficiency, water use, and waste heat recovery. Draft criteria have already drawn criticism from industry and some Member States, especially over concerns that nuclear energy could be treated less favorably than renewables. Jørgensen indicated that the Commission intends to reflect technology neutrality, an important point for lawyers advising on EU energy, digital infrastructure, and sustainability rules.
A key regulatory concern is transparency. Current reporting under EU energy-efficiency rules remains incomplete, with only a minority of covered data centers reportedly submitting required data. The Commission also wants operators to make better use of waste heat, which could help heat millions of European homes. For data center developers and AI companies, the legal message is clear: market access and public acceptance in the EU will increasingly depend on credible clean energy sourcing, local grid integration, heat reuse, and measurable sustainability performance.
