Original article Total Telecom:
Feature Week
At this year’s inaugural Data Centre Summit, co-hosted alongside the Connected North conference in Manchester, discussions focussed heavily on the government’s emerging strategy around digital infrastructure and AI.
In September last year, the UK government announced that it would officially class data centres as critical national infrastructure, ensuring they receive heightened levels of protection against cyber threats and IT disruption.
The decision, announced by Technology Secretary Peter Kyle, “will allow better coordination and cooperation with the government against cyber criminals and unexpected events.”
The classification, which has long been advocated by industry leaders, elevates data centres to the same level of importance as water, electricity, and gas.
“To call it critical is probably underplaying it because the minute you lose it, the world knows about it,” said Mark Yeele, Vice President of Security at Schneider Electric, speaking to Total Telecom at last week’s Connected North event. “I’ve been advocating for [this classification] since I’ve been in the job. It’s really the fourth utility.”
This designation marked a growing government appreciation for the integral role that data centres play in the country’s digital life, as well as their paramount importance in enabling an AI-powered economy.
“Cloud computing, the internet, and now AI all lives in the physical data centre, in the physical building. And these physical buildings are the foundation of our technological lives. Without them, we won’t be able to advance forward and be aggressive in the AI world,” said Spencer Lamb, Chief Commercial Officer at KAO Data.
Leveraging AI Growth Zones
At the heart of the government’s approach to AI is the proposed creation of AI Growth Zones. These Zones are aimed at accelerating AI data centre deployment improving access to power and providing planning support.
Exactly what form this support will take it unclear; the government is currently seeking input from regional and local authorities, as well as the data centre and power industries themselves.
What is clear, however, is the potential this initiative has for the North of the UK. Designed to stimulate investment in AI infrastructure and create high-skilled jobs, AI Growth Zones are especially focused on deindustrialised and economically underpowered regions, many of which are in the North. Areas such as Greater Manchester, South Yorkshire, and Teesside are strong contenders to become AI hubs due to their existing universities, digital clusters, and ambitions to drive tech-led growth.
“We see the AI Growth Zones as much more than a data centre building opportunity,” said John Duncan, Connected Places Lead – Greater Manchester Combined Authority. “We don’t really know fully what they will look like yet, but what I have liked is that the government are interested in talking to regional and local authorities about where they will see the benefits of data centre deployments. We’ve had data centre operators talking to us not only AI training and R&D, but also how it’s going to support public sector services and local businesses.”
These collaborative discussions represent a significant shift in mentality for the data centre industry, which have typically been very siloed and largely unregulated. The launch of the AI growth zones is bringing industry stakeholders closer together. Operators used to focus inwardly on building and maintaining their own sites, rarely interacting beyond commercial partnerships or local planning permissions. There was no organised public voice for the sector, no formal representation at government level, and little need to engage with wider national strategy. The UK government has put the sector firmly on the national agenda, meaning that for the first time, cross industry collaboration is essential.
Training or inference?
But underpinning these discussions is a larger question: what role does the UK want to play in the global AI ecosystem? Does it want to become an AI training hub or instead focus merely on inference? The infrastructure requirements for each are completely different.
The infrastructure requirements for each are quite different. Training requires vast amounts of power, high-density data centre design, advanced cooling systems, and proximity to large-scale compute clusters. Inference, by contrast, is more distributed and can run on lighter infrastructure, closer to the edge or within enterprise settings.
“The Department of Science and Information and Technology actually taking it seriously means we now have got their ear and we can encourage them,” said Spencer Lamb. Do you want the UK to become an AI training enclave or do you just want to do inferencing? The answers to those questions are very different.”
Data centre regulation: A Brexit benefit?
These decisions could have major ramifications for the UK’s technological standing on the international stage. Part of Europe but at the same time separate from it, the UK’s unique post-Brexit position could make it something of an AI gateway, taking the best parts of EU regulation and rejecting those that slow development.
“Ideally, I’d like to see it become an AI training country,” said Lamb. “Our independence from the EU […] means we can do our own law changes around data centres without being governed by the unnecessary bureaucracy. There’s an opportunity here, supported by our many renewable energy schemes.”
“If the UK government do it the right way then I think this is the catalyst to put in the UK seriously on the map as an AI idea,” added Karl Havard, Chief Commercial Officer at Nscale.
Beyond regulation, sovereignty over cloud and AI infrastructure is becoming a national priority. Control over where data resides, and the laws that govern it, will be crucial for the UK’s long-term digital resilience.
“There is a need now for countries across the EU and the UK to increasingly to be in control of our own AI architecture, not just its location but actually the laws that govern it,” Havard explained. “That leads to the big question of why is it currently that all the hyperscalers are from the US? Why can’t we create our own?”
“I think we can help the UK government to see that it’s a bigger risk to [rely on major US cloud providers] rather than actually creating something that is in control of the UK itself. I’m actually very hopeful that we can get to a position where we start to see the UK building its own domestic hyperscalers,” he added.
A new focus
The designation of data centres as critical infrastructure, and the rollout of AI Growth Zones mark a turning point, and the optimism in the industry was clear to see at Connected North. But with rising power costs, growing demand, and a fragmented legacy industry, the UK’s ability to deliver on its AI potential will depend on whether it can build and govern the infrastructure to match.
