The European Commission has opened a new consultation that could shape the future of data centers across Europe.

Launched on September 21, 2026, the consultation seeks feedback on minimum performance standards for data centers. Stakeholders have until December 14, 2026, to respond.

A legislative proposal is currently planned for Q2 2027.

Why is the EU consulting now?

Data centers are becoming increasingly important to Europe's digital economy.

The growth of AI is accelerating that demand.

But data centers consume significant amounts of electricity and water, while also creating pressure on electricity grids and other resources.

The European Commission says it wants to balance digital expansion with energy efficiency, resilience and sustainability.

What could change?

The consultation will help determine which indicators should have minimum performance standards and what thresholds would be appropriate.

The EU is already developing a common rating system for data centers above 500 kW, covering resource use and interaction with electricity grids.

Possible future standards could address:

Energy efficiency

Water consumption

Waste heat

Grid flexibility

Sustainability

Resource utilization

Why should AI companies care?

AI infrastructure is physically intensive.

Every AI query ultimately requires computing resources somewhere.

For organizations running large AI workloads, infrastructure efficiency is becoming a business issue as well as an environmental one.

Companies may increasingly need to consider where their workloads run, how much energy they consume and whether infrastructure providers meet emerging standards.

What happens next?

The current consultation does not establish final requirements.

It is part of the evidence-gathering process before potential legislation.

But businesses operating data centers or relying heavily on AI infrastructure should pay attention.

Europe's AI policy conversation is expanding.

It is no longer only about models and data.

It is increasingly about the physical infrastructure required to run AI at scale.