Energy Costs Remain Out of Sight in the AI Act
As the EU’s AI Act takes effect, a key question looms: can Europe balance AI innovation with its growing energy demands?
AI is transforming the energy sector, offering unprecedented opportunities to increase efficiency, integrate renewables, and optimise electricity grids. However, behind this promise lies a growing challenge: AI’s energy consumption is surging, driven largely by the expansion of data centers. According to the International Energy Agency, data centres account for 1.5% of global electricity use today, a figure that is expected to more than double by 2030. In the EU, they already represent about 3% of total demand, with stark regional disparities.
Data centres account for 1.5% of global electricity use today, a figure that is expected to more than double by 2030.
The EU has set climate neutrality targets for data centres by 2030, alongside reporting obligations on energy use, water consumption, and renewables. But the rapid growth of AI-focused facilities is putting pressure on local grids and testing the balance between digital ambition and environmental sustainability. A single AI query can consume up to 10 times the electricity of a standard Google search, especially for multimedia-rich outputs.
The act goes quiet on sustainability
The EU AI Act, finalised in 2024 and entered its implementation phase in August 2025, is primarily known for its risk-based regulatory framework for AI systems. However, the law’s intersection with the energy sector is emerging as a critical and contested dimension. While the AI Act includes transparency and risk management obligations for high impact AI systems, it doesn’t impose direct sustainability requirements yet.
Some argue that without binding efficiency and emissions provisions in the AI Act itself, Europe risks fueling an unsustainable AI boom. Others counter that the AI Act is a horizontal regulation and that energy-specific measures should remain under separate instruments, such as the planned ‘data centre energy efficiency package’ in 2026.
The political fault line mirrors other EU debates: some see sustainability clauses as essential for aligning AI development with the EU’s climate goals, while others warn that adding such requirements could slow innovation and burden small providers.
What’s next for AI and the energy sector?
Over the rest of 2025, the European Commission will decide how tightly to link the AI Act’s implementation to energy-efficiency policies.
Two pathways are on the table:
- Integration path: Introduce delegated acts or accompanying guidance under the AI Act that require high-energy AI systems and data centers to meet sustainability benchmarks.
- Separation path: Keep the AI Act focused on safety, transparency, and ethics, leaving energy concerns to the forthcoming sector-specific legislation.
Either choice will shape how Europe navigates the trade-off between digital competitiveness and climate responsibility. A narrow focus risks creating a regulatory gap, while a coordinated approach could strengthen Europe’s leadership in both AI governance and sustainable innovation.
Ultimately, this debate goes far beyond the technicalities of algorithm oversight. It is a test of whether the EU can expand AI’s potential while staying true to its climate-neutral ambitions, or whether short-term political and economic priorities will once again take precedence over environmental responsibility.