The debate is becoming less about attracting investment and more about identifying which projects can actually be delivered
Last year was the year of AI announcements. AI Growth Zones, sovereign AI, multi-billion-pound pledges and new data centre campuses, pledges that all show Britain is serious about competing for the infrastructure behind the digital economy.
That is not abstract. Data centres sit behind many of the services people use every day, whether its streaming Netflix, storing photos in the cloud, making card payments, using online banking, booking GP appointments, navigating with Google Maps, working on Teams and accessing public services online. They are also becoming increasingly important to the AI tools businesses and public services are beginning to rely on. The communications challenge for the sector is to tell that story clearly, rather than assuming communities will accept a project simply because it comes with a large investment number attached.
The Government, itself, is changing too – recent machinery of government changes have moved parts of the old science and technology brief into a new expanded business department. Kanishka Narayan now leads on AI and attends Cabinet, while Lord Vallance chairs the Prime Minister’s AI Taskforce. The ambition is there, but the question is whether the infrastructure can keep pace.
For much of the past two years, the debate focused on attracting investment. That made sense. Britain wanted to show that it was open for business and serious about AI.
But the UK is no longer short of proposed data centre developments. Instead, government, investors and communities are asking harder questions about which projects can actually be built, how they will be powered and what benefits they will leave behind.
The conversation is moving from why data centres matter to how they get delivered.
The biggest challenge facing the sector is not interest, but power. Data centres are energy intensive and operate around the clock. As demand for AI grows, so too will demand for electricity. At the same time, projects in the UK face long waits for grid connections, while recent reporting on delayed projects has shown how quickly power access can become the defining risk.
That raises a basic question: where does the power come from?
Projects relying on future grid reinforcement may be in a different position to those already thinking seriously about energy supply, storage and resilience. Increasingly, developers are exploring models that combine data centres with generation, batteries and other energy infrastructure.
That could include Energy from Waste facilities, which offer stable local power and are beginning to be viewed not just as waste assets, but as partners in the UK’s digital economy.
The exact solution will differ from project to project, but energy is no longer a supporting consideration. It is becoming one of the factors that determines whether a project can move from proposal to delivery.
Concerns about proposed hyperscale developments have led to discussions about a potential pause on new applications while policymakers better understand capacity and system impacts.
Governments are becoming less interested in the headline scale of a proposal and more interested in credibility. A data centre campus promising hundreds of megawatts may look impressive, but if questions remain around how it will be powered, funded, consented and accepted locally, its path becomes less certain.
By contrast, a smaller project with a clear energy strategy, realistic timeline and strong local support may prove more valuable.
Readiness, credibility and a clear route to delivery are becoming increasingly important. Projects that can answer questions on power, planning and local benefit are likely to move to the front of the queue.
The original idea behind AI Growth Zones was sensible: identify places with the right conditions and help remove barriers to investment.
That logic is still strong. If the UK wants to compete for AI infrastructure, it needs a mechanism that brings together planning, power, local growth and private investment.
But recent Whitehall changes mean government needs to be clear about who owns the programme and how decisions will be made. With DSIT folded into the new expanded business department, and AI given more prominence through Kanishka Narayan and Lord Vallance, the direction of travel is clearly not away from AI.
The test now is whether AI Growth Zones can help serious projects move from proposal to delivery.
That means prioritising schemes with credible funding, realistic timelines, clear power arrangements and a strong local benefits case. If Growth Zones can do that, they could become an important part of the UK’s industrial strategy.
The UK will need more digital infrastructure if it wants to compete in AI, strengthen digital resilience and support future economic growth.
But the projects that attract the most support are likely to be those that can demonstrate clear funding, realistic timelines, credible energy solutions and tangible benefits for host communities.
The UK does not have a shortage of ambition. What it needs now is a pipeline of projects that can turn that ambition into reality.
Britain’s AI future will not be determined by announcements alone, but by the infrastructure that gets built, the services it enables and the value it creates.
Blog
