Debate is raging about whether the UK government’s push to grow the number of homegrown datacentres with capabilities to host compute-intensive artificial intelligence (AI) will accelerate the adoption of AI technologies across public and private sectors.
As reported by Computer Weekly, the government has published its AI opportunities action plan policy paper, which details 50 actions the government has committed to delivering on to ensure the use of AI technologies and services becomes more pervasive across the UK.
According to the government, the plan means “AI will be unleashed across the UK to deliver a decade of national renewal” by enabling the technology to take on more of the public sector’s administrative burden.
This, in turn, will free up time for employees to concentrate more on delivering frontline services that benefit citizens directly, with the government citing the NHS as a use case where this type of work is already reaping benefits.
“[AI] is being used in hospitals up and down the country to deliver better, faster and smarter care: spotting pain levels for people who can’t speak, diagnosing breast cancer quicker, and getting people discharged quicker,” the government said in a statement. “This is already helping deliver the government’s mission to build an NHS fit for the future.”
Establishing infrastructure for AI adoption
Accelerating the take-up of AI technologies and services will require sizeable investments in building out the UK’s AI-ready datacentre infrastructure.
This is something the government has wasted no time in encouraging operators to get involved with since coming to power in July 2024, and it has used the policy paper’s publication to name-check several operators that have recently committed to building server farms in the UK.
They include Vantage Data Centres, which is in the throes of building a large datacentre campus in Wales and has committed to invest £12bn in building server farms across the UK, and NScale, which has also vowed to invest £2bn in building AI datacentres in the UK over the next three years.
Incidentally, the government also confirmed that more than £25bn in private sector investment in UK datacentres has been secured since it was voted in.
However, more supporting infrastructure will be required, which is why one of the standout parts of the AI action plan is a commitment to expand the UK’s sovereign compute capacity at least 20-fold by 2030.
To achieve this, the government has committed to creating a supercomputing facility that it claims will “double the capacity” of the UK’s AI research resource. The Department for Science, Innovation and Technology (DSIT) is already building a business case and sourcing suppliers for the facility, the government confirmed in the policy paper.
“The UK will also work with stakeholders to develop a long-term compute strategy that will ensure the UK has the AI infrastructure and compute capacity it needs to deliver new scientific innovations and discoveries that will drive productivity and growth throughout the economy,” stated the AI opportunities action plan: Government response policy paper.
With the contents of the AI action plan now public, the pressure is on the government to turn its rhetoric into reality and tangible economic results
The paper also goes into more detail about the government’s plan to create a series of AI growth zones, which will be designated areas of the UK that have “enhanced access to power and support for planning approvals”.
This aligns with promises the government has previously made about lowering planning permission barriers for datacentre developments in the UK.
As detailed in the policy paper, the first pilot AI growth zone is on course to be built in Culham, at the headquarters of the UK Atomic Energy Authority (UKAEA), which will work with the government from spring 2025 to recruit a private sector partner to deliver “one of the UK’s largest AI datacentres” on the site.
“The government and UKAEA will seek a private sector partner who would develop one of the UK’s largest AI datacentres, beginning with 100MW of capacity and with plans to scale up to 500MW,” the paper stated.
“The pilot would pioneer innovative public-private models to deliver secure, dedicated public sector computing capacity, supporting key national priorities.”
While the pilot AI growth zone project progresses, efforts will be made to identify other areas of the UK that might make suitable additional sites for this initiative, the paper added.
“AI growth zones have the potential to grow the AI sector to support AI adoption across the economy and enable the government to build new strategic partnerships with leading AI players,” the policy paper continued.
“They also have the potential to deliver local benefits, including job creation, enhanced digital and energy infrastructure and sustainability initiatives.”
Industry feedback
With the contents of the AI opportunities action plan now public, the pressure is on the government to turn its rhetoric into reality and tangible economic results. But questions are being asked about whether or not the government is up to the task.
Simon Baxter, principal analyst at IT analyst house TechMarketView, pointed out – in an email to the company’s subscriber base – that creating more datacentres does not necessarily mean that AI usage will increase in the UK public sector.
“Certainly, when it comes to the use of AI in public services, there are no clear timelines on how quickly the technology can be implemented, and there is a risk that the impact of AI may be far less than expected,” he said.
When it comes to the use of AI in public services, there are no clear timelines on how quickly the technology can be implemented, and there is a risk that the impact of AI may be far less than expected Simon Baxter, TechMarketView
At the same time, with public sector organisations facing ever-tightening financial constraints, the pressure will be on public sector bodies to achieve “early gains” from adopting AI. “And if they don’t, they will struggle to continue investing when fiscal times are tougher,” Baxter continued.
“We are also still missing enough clear examples of the returns public departments are achieving from AI deployments, which would be a far more useful metric than these announcements of numerous AI activities.”
Mark Boost, CEO of UK-based cloud services provider Civo, is far more optimistic about the impact the government’s plans will have on AI usage among more risk-averse users within the public sector, on the basis that the action plan is championing the creation of sovereign AI services.
“Data sovereignty means that all data is processed and stored within our borders, and is subject only to UK laws,” said Boost. “With this guarantee, public and private sector organisations in healthcare, defence, law and many more will finally be able to deploy AI without fear of breaching data protection controls.”
John Buyers, global head of AI at legal firm Osborne Clark, told Computer Weekly of his concerns about how much “overarching responsibility” the plans put on the government and “bureaucratic mechanisms” to accelerate the growth of the UK’s AI sector.
“Are these the right mechanisms for such a fundamental initiative or should great reliance have been placed on nurturing the private sector?” he said. “Looking to the public sector to lead on innovation, especially on large-scale infrastructure projects, is not something that the UK has a happy history of.”
Buyers also raised questions about where funding for the sovereign compute part of the AI opportunities action plan is going to come from, given the expense involved in building datacentres.
“Large AI providers in the United States are spending billions of dollars on a quarterly basis to fund datacentre and energy commitments – including recommissioning nuclear power stations for extra energy,” said Buyers. “Levels of taxation are already at record levels – and tax is the only capital revenue source that government can rely on.”