← All insights
Infrastructure and cloud Sustainability and reputation

The cloud has a postcode

AI infrastructure is colliding with grids, planning systems and local communities. The next big technology story will not be won with scale alone; it needs a credible account of place and impact.

Written by David Carter
3 min read Share on LinkedIn ↗
Rows of illuminated servers in a modern data centre
Photo: Taylor Vick / Unsplash

“The cloud” was a magnificent piece of branding. Weightless, clean and conveniently located somewhere above the difficult questions.

Unfortunately, AI has given the cloud a postcode.

Data centres need land, grid connections, cooling, equipment, planning consent and neighbours who may reasonably ask what they get in return. The International Energy Agency says global data-centre electricity consumption grew 17% in 2025 and could rise from 485 TWh that year to roughly 950 TWh in 2030. It also notes growing community concern about affordability and environmental impact.

This does not make data centres the villain of the decade. It makes “invisible infrastructure” an obsolete communications strategy.

For operators, cloud providers, utilities and the businesses building around them, the story can no longer begin and end with investment, jobs and innovation. Those claims may be true, but they are the standard opening lines of every infrastructure announcement. Communities and decision-makers need the next page.

Start with the postcode

Start with place. Why this location? What existing constraints shaped the choice? How will the project interact with local power, water, transport and skills? A national economic argument will not answer a local concern.

Then make the trade-offs legible. Sustainability language often fails because it is either too polished to trust or too technical to understand. “Powered by renewable energy” invites obvious follow-up questions: which source, at what times, through what agreement and with what effect on the wider system? Good communication does not hide complexity. It organises it.

Use numbers with context. A large figure without a comparison creates drama, not understanding. A percentage without a baseline creates suspicion. Show the measure, the period, the source and what remains uncertain. If the modelling changes, say so before somebody else does.

Communicate before the planning notice

Most importantly, communicate before the planning notice appears. Stakeholder engagement is not a presentation delivered after the important decisions have been made. Listen early enough for the answers to affect the plan. Publish the questions you heard. Explain what changed and what did not.

The visual language matters as much as the statement. A glossy render of a silver building beneath a spotless blue sky says very little about grid pressure or local benefit. Use maps, timelines and simple diagrams to show construction, connections, mitigation and accountability. Good information design lets a resident, councillor or investor interrogate the same evidence without forcing all three through an engineering report. It also makes uncertainty visible: planned, contracted and operational capacity are not interchangeable, however tempting the headline may be.

There is a wider opportunity here. Energy is not simply a defensive topic for AI businesses. Efficiency, grid flexibility, better forecasting and smarter industrial systems are part of the technology’s value. But that positive case only lands when it sits beside, rather than on top of, the physical cost.

The strongest infrastructure narratives will connect three things: what the technology enables, what the place experiences and how the company will be held accountable. That requires strategy, public relations, design and local engagement to work as one system.

The cloud was never weightless. Now that everyone can see the cables, the story needs to carry its share of the load.

From abstraction to infrastructure

Make the system legible.

Maps, diagrams and clear visual evidence help communities, decision-makers and investors understand what infrastructure means for a place.

Abstract data services and digital infrastructure visual
Data and infrastructure need context people can interrogate.

Sources and further reading

  1. IEA Key Questions on Energy and AI executive summary ↗
  2. IEA Energy and AI analysis ↗

Where this becomes useful

Turn infrastructure into understanding

Talk to us about your next story