The UK government is pushing hard to increase the number of datacentres that can support heavy-duty artificial intelligence (AI). The big question is whether this will actually speed up AI adoption in both public and private sectors. Recently, the government released its AI opportunities action plan, laying out 50 actions aimed at making AI more common across the country.
The plan suggests that AI can help ease some of the administrative workloads in public services. For instance, the NHS is already seeing benefits from AI usage, like identifying pain levels in non-verbal patients and speeding up breast cancer diagnoses. The government believes that by automating these tasks, healthcare workers can focus more on services that directly help citizens.
To make this happen, the UK needs a robust datacentre infrastructure. Since taking office in July 2024, the government has urged operators to develop AI-ready facilities. Companies like Vantage Data Centres are investing £12 billion in building datacentres across the UK, while NScale has committed to £2 billion for AI datacentres over the next three years. Since the current government came into power, over £25 billion in private investment for datacentres has already been secured.
However, the government aims to expand the UK’s computing capacity 20 times by 2030. One pivotal part of the plan is to create a supercomputing facility that will double the UK’s AI research capability. The Department for Science, Innovation and Technology (DSIT) is actively working on this, researching suppliers to build this facility.
The government is also focusing on creating AI growth zones—areas with improved access to power and expedited planning approvals for datacentre projects. The first such zone will be in Culham, at the UK Atomic Energy Authority’s headquarters, where a massive AI datacentre is set to be built. This facility aims to deliver 100MW of capacity initially, with plans to scale up to 500MW, highlighting a public-private partnership that supports national priorities.
Yet, industry experts have mixed feelings. Simon Baxter from TechMarketView cautions that simply increasing datacentres doesn’t guarantee greater AI use in public services. There are no clear timelines on implementation, and he worries that outcomes might not meet expectations. With financial constraints in the public sector, organizations will need to show early benefits from AI or risk losing funding.
On a more positive note, Mark Boost, CEO of Civo, believes that the plans could encourage more public sector organizations, especially those that are hesitant, to use AI because of data sovereignty. This ensures that data is processed under UK laws, giving organizations in sectors like healthcare and defense the confidence to deploy AI without risking data breaches.
However, there are concerns about the role of the government. John Buyers from Osborne Clark questions whether relying on government mechanisms is the best way to innovate in the AI space. He raises doubts about funding for the ambitious computing capacity goals, especially as large AI firms in the U.S. invest billions into their datacentre infrastructure, and highlights the heavy taxation that already exists in the UK.
With expectations high, the focus now is on whether the government can deliver on these ambitious promises and make meaningful advancements in the AI sector.