# Europe must demand proof before powering AI data centres

> Europe should require AI data centres to demonstrate financial viability and public benefit before granting grid priority, not treat them as automatic infrastructure needs.
- Title: Europe must demand proof before powering AI data centres
- Summary: Europe should require AI data centres to demonstrate financial viability and public benefit before granting grid priority, not treat them as automatic…
- Keywords: artificial intelligence, data centres, electricity grid, europe, infrastructure, energy policy, grid capacity, manufacturing, technology, startups, must, demand
- Source: The Next Web — https://thenextweb.com/news/europe-ai-data-centres-power-grid-priority
- Author: Paul M. Sotkiewicz
- Published: 2026-10-28T17:41:02+00:00
- Read time: 2 min
- Topics: artificial intelligence, data centres, electricity grid, europe, infrastructure, energy policy, grid capacity, manufacturing, technology, startups
## Who pays for AI's power needs
When data centres connect to electricity grids, questions arise about who pays for supporting infrastructure and what other opportunities that investment could serve.
## Data centre power demand doubling
The International Energy Agency projected global data centre electricity consumption would reach around nine hundred forty five terawatt-hours by twenty thirty, more than twice its twenty twenty four level.
## AI pilots show no returns
An MIT study found that ninety five percent of companies' generative AI pilots had produced no measurable return, and the economics of AI data centres have not been shown to be profitable.

> 95% of companies' generative AI pilots had produced no measurable return
## Developers must prove demand first
Data centres and AI developers seeking scarce grid capacity should demonstrate credible demand, accept the costs they create, and explain public benefits when requesting approvals.
## US lesson: demand shrinks with costs
AEP Ohio reported an initial pool of roughly thirty gigawatts of data centre requests narrowed to about thirteen gigawatts after customers had to request and pay for engineering studies.
## Binding agreements cover far less
Binding agreements from AEP Ohio's pool subsequently covered approximately five point six gigawatts, excluding contracts signed before the new tariff, showing how financial commitment filters demand.
## Manufacturing and security at stake
Europe faces a generational challenge regarding industry capacity and regional security. Manufacturing, domestic energy equipment production, and defence production must not be sacrificed for unproven AI data centre value.
## Infrastructure timelines don't align
Data centres become operational in two to three years, but energy infrastructure requires longer. Natural gas turbines wait five years, step-up transformers three years, creating near-term constraints.
## Regulators should require commitments
Regulators should require credible development milestones and financial commitments before incorporating large proposed loads into infrastructure plans, with developers paying grid upgrade costs attributable to their projects.
## AI must earn its place
European policymakers should make proof-of-value requirements part of connection rules and public funding decisions now. AI should earn its place through credible commitments and demonstrable value.

> AI should earn its place in the energy system through credible commitments and demonstrable value
## Key takeaway

Europe should require AI data centres to prove credible demand and profitability before prioritising them for scarce grid capacity, rather than assuming their value.