An On the wire story
Energy Minister Michael Shanks has said the UK must ensure the build-out of artificial intelligence infrastructure translates into lower household bills, warning that public support could fade if data centres are seen simply as another burden on an already strained power system.
Speaking at the UKAI Infrastructure Summit, he argued that the government has to show AI can strengthen energy security and improve affordability, not just consume scarce electricity capacity.
His intervention comes as Britain’s AI ambitions run into a practical obstacle: the grid. City AM reported that Nscale is exploring alternative power arrangements for its planned £2bn Essex data centre after delays to a National Grid connection, while more than 100 UK developments are said to be weighing on-site gas generation while waiting for access to the network.
Some projects are facing connection waits stretching close to a decade, even as ministers want data centres to anchor domestic AI growth.
The government has also revised sharply upwards its estimate of the carbon footprint linked to its AI infrastructure plans. According to the Department for Science, Innovation and Technology, AI data centres could emit between 34 million and 123 million tonnes of carbon dioxide over the next decade, depending on efficiency gains and how quickly the power system decarbonises.
The upper end of that range would amount to about 3.4 per cent of Britain’s projected emissions between 2025 and 2035.
Shanks said one answer would be to site data centres more strategically, particularly near renewable generation that is currently curtailed when the network cannot carry power south from Scotland. By soaking up excess electricity that would otherwise be wasted, he said, such facilities could help reduce bills and spread the cost of the system across a broader base of industrial users.
The minister also suggested AI could make better use of the existing network, reducing dependence on new pylons and substations, while industry voices at the summit pointed to solar and battery storage as ways for households to manage costs and improve resilience. Researchers and policymakers are increasingly focused on the same tension: how to expand AI capacity without undermining climate targets, pushing up prices or overloading the grid.
References
• This On the wire article was created using Noahwire AI and reviewed by CIBSE Journal’s editorial team
