State of AI for Decarbonisation 2025
Abstract AI has been constantly in the headlines this year. Jaw-droppingly large investments in AI data centres underpinned US economic growth. Securing enough energy for those data centres became a major concern and tech companies started funding nuclear power plants and grid reinforcement. Businesses around the world started integrating generative AI into their software and […]
Abstract
AI has been constantly in the headlines this year. Jaw-droppingly large investments in AI data centres underpinned US economic growth. Securing enough energy for those data centres became a major concern and tech companies started funding nuclear power plants and grid reinforcement. Businesses around the world started integrating generative AI into their software and processes, with mixed results. Debates about the future of work intensified, with people either worried or excited about AI’s potential to automate jobs. Even national carbon budgets started including assumptions that AI could save millions of tonnes of CO2 emissions. But amidst that torrent of news and hype it can be hard to find the answer to an important question: how effectively is AI being applied to key societal challenges like decarbonisation?
This annual report answers that question with a review of how UK applications of AI for decarbonisation have matured over the last year. This ranges from applications that are reaching scale and meaningfully contributing to national decarbonisation, through to earlier stage research that has made notable progress. There have been tangible steps forward in some areas this year. AI-powered EV charging is already playing a significant role in managing our low carbon electricity grid. Heat pump installations are quicker and cheaper because of AI. Steel furnaces and cement plants are reducing emissions through AI-optimised operations. But in other areas progress has been slow or hindered by generative AI hype. Lots more work is required to fully realise the benefits of AI for decarbonisation.
Citation information
ADViCE. (2025). State of AI for Decarbonisation 2025. https://www.turing.ac.uk/research/research-projects/advice.
Additional information
The AI for Decarbonisation Virtual Centre of Excellence (ADViCE) is a collaboration between Digital Catapult, Energy Systems Catapult and The Alan Turing Institute, funded by the Department for Energy Security and Net Zero.
The report was officially launched via a webinar in January 2026. For those interested in a guided overview of the key findings and discussion points, a recording of the session is available here.
Last updated: