AI Growth Zones are ‘a policy decision without a business case’
The operator, which designs, builds and operates data centres in Europe, the Middle East and Africa, welcomed the recognition of data centres as critical national infrastructure. But a senior source at the company said AI training data centres relied heavily on sources of cheap electricity, while developers of inference data centres required locations near urban centres.
“The challenge is that the AI Growth Zones are a policy decision without a business case,” they said. “There was no consultation prior to that policy being launched. Power pricing is where it comes unstuck. AI training zones need cheap power. If our power is four times more expensive than the US, nobody is going to come here.”
Inference data centres, meanwhile, needed to be located close to cities because of latency. “The problem is these AI growth zones are up in the north of Scotland, or in the Northeast of England. They aren’t close to major centres where enterprises want to locate inference data centres, so that business case is a busted flush.
“Who is actually going to occupy these data centre locations?”
Connectivity to the grid also posed a challenge. “There’s an issue with power transmission and substation development. We’ve heard of developers being quoted 10 or 20 years before they get a connection.
“The AI bubble will have burst by then and all investment moved overseas.”
The Department for Science, Innovation and Technology denied that the zones were being established without consulting industry. A spokesperson said: “The recent package of policy announcements we have confirmed reflects the many conversations we've had with AI Growth Zone applicants and wider industry regarding concerns over energy pricing and grid connections.”
The department added that the government's AI Energy Council – co-chaired by the Secretaries of State for Technology and Energy Security and Net Zero – had been established to focus on sustainable and innovative solutions to power the UK's AI infrastructure. At a meeting in early December, the council discussed energy grid reforms to help improve energy access and deliver greater certainty.
The government has also announced a package of reforms to accelerate delivery of the next generation of AI data centres through the AI Growth Zones Programme.
“This includes measures for data centres in AI Growth Zones to get significant discounts on their electricity bills, if they are located in areas that could help to reduce the pressures on the energy network,” the spokesperson said.
“Developers could also be supported to connect their own high voltage lines and substations to power their data centres – rather than waiting for network operators to do it – driving down costs and accelerating progress.”
Further announcements on future AI Growth Zones are expected.
“The challenge is that the AI Growth Zones are a policy decision without a business case,” they said. “There was no consultation prior to that policy being launched. Power pricing is where it comes unstuck. AI training zones need cheap power. If our power is four times more expensive than the US, nobody is going to come here.”
Inference data centres, meanwhile, needed to be located close to cities because of latency. “The problem is these AI growth zones are up in the north of Scotland, or in the Northeast of England. They aren’t close to major centres where enterprises want to locate inference data centres, so that business case is a busted flush.
“Who is actually going to occupy these data centre locations?”
Connectivity to the grid also posed a challenge. “There’s an issue with power transmission and substation development. We’ve heard of developers being quoted 10 or 20 years before they get a connection.
“The AI bubble will have burst by then and all investment moved overseas.”
The Department for Science, Innovation and Technology denied that the zones were being established without consulting industry. A spokesperson said: “The recent package of policy announcements we have confirmed reflects the many conversations we've had with AI Growth Zone applicants and wider industry regarding concerns over energy pricing and grid connections.”
The department added that the government's AI Energy Council – co-chaired by the Secretaries of State for Technology and Energy Security and Net Zero – had been established to focus on sustainable and innovative solutions to power the UK's AI infrastructure. At a meeting in early December, the council discussed energy grid reforms to help improve energy access and deliver greater certainty.
The government has also announced a package of reforms to accelerate delivery of the next generation of AI data centres through the AI Growth Zones Programme.
“This includes measures for data centres in AI Growth Zones to get significant discounts on their electricity bills, if they are located in areas that could help to reduce the pressures on the energy network,” the spokesperson said.
“Developers could also be supported to connect their own high voltage lines and substations to power their data centres – rather than waiting for network operators to do it – driving down costs and accelerating progress.”
Further announcements on future AI Growth Zones are expected.
