UK departments at odds over energy demands of AI datacentres | AI (artificial intelligence)
One vision for the UK’s future includes a decarbonised economy powered by clean, renewable energy. Another involves turning the UK into an AI superpower.
The government departments responsible for these two visions do not seem to agree on their numbers.
The Department of Science, Innovation and Technology (DSIT) thinks AI data centers will consume 6 GW of electricity by 2030. The Department for Energy Security and Net Zero (DESNZ) seems to think they will use less than a tenth of that.
Tim Squirrell, head of strategy at NGO Foxglove, said: “The government’s ignorance of the environmental impact of data centers would be laughable if it weren’t so alarming.”
Researcher Cecilia Rikap, of University College London, said: “There are two ways to interpret this ‘misalignment’: either DESNZ and DSIT are incompetent, or there is some magical thinking about AI and big tech. Either way, this episode reveals how these companies control not only the AI value chain but also the UK government.”
DESNZ is responsible for the UK’s carbon budget growth and delivery plan, which sets out how the government will meet international climate targets.
In January, Foxglove asked the department for an environmental impact assessment, asking how it had factored AI data centers into the UK’s emissions estimates. In response, DESNZ referred researchers to its broader estimates of energy use by Britain’s “business services” sector and said it did not make separate estimates of data center growth.
Forecasts appear to predict that the energy use of the entire sector will increase by 528 MW between 2025 and 2030; This is equivalent to adding the consumption of 1.7 million homes by the end of the decade.
This projection is 10 times lower than the amount of electricity the government owns. loyalty as part of artificial intelligence data centers UK computing roadmap. This policy document, put forward by DSIT in 2025, lays out a “bold, long-term plan to transform our national computing ecosystem” by creating AI data centres.
He adds: “We estimate the UK will need at least 6GW of AI-enabled data center capacity by 2030.”
This will come from multiple AI growth zones, hubs across the country where the government is trying to attract investment in data centers. Each will require at least 500 MW of electricity; this amount is only slightly less than DESNZ’s estimate of the increase in energy use of the entire business services sector.
It is unclear how the difference between the two ministries’ estimates arose. But a day after the Guardian requested comment from DSIT and DESNZ, DSIT appears to have revised the AI data center sector’s total emissions figures published on its website by more than a hundredfold.
Initially, DSIT’s estimates of carbon emissions from additional AI computing capacity were between 0.025 million tonnes and 0.142 million tonnes carbon equivalent (MtCO₂); this was below 0.05% of Britain’s projected emissions. These figures were included in the appendix of the IT roadmap.
Earlier this year this document was removed from the government website After Carbon Brief raised questions on the matter possibility from numbers. DSIT updated its figures on Thursday after the Guardian asked about it.
In a statement published online, in question: “The UK’s cumulative 10-year greenhouse gas emissions from AI calculation could range from 34 to 123 MtCO₂, which is approximately 0.9-3.4% of the UK’s total projected emissions over the 10-year period.
“If successful, the UK’s grid decarbonisation plans will help reduce emissions from data centers towards the lower end of this range,” he added.
A DESNZ spokesperson said: “Data center emissions are included in our modelling, including carbon budget 7” and added: “The AI Energy Council is exploring opportunities to attract investment and support the development of clean energy for data centres.”
Carbon budget 7 will be published this summer. A DSIT spokesman referred the Guardian to DESNZ.
