Green AI – How to Make AI Sustainable

Green AI and green energy: twin transitions

Leaders who seize the opportunity to support green AI across its whole value chain can foster a positive loop in which the twin transitions of AI and energy work in synergy to deliver benefits. AI can help speed up the scientific discoveries and innovations needed to move towards cleaner sources of energy, while the shift to clean energy can improve efficiency and lower costs – all of which can create ideal conditions for further technological innovation. Green AI will give countries an opportunity to grow their information and communication technology, make the most of their computing power while still meeting climate goals and attract major tech companies looking for new investment locations.

Media coverage of our green-AI expertise

On 16 September 2025, a BBC News article discussing AI investments in the UK as part of the “Tech Prosperity Deal” between the US and the UK quoted Keegan McBride on the next steps required for the UK to fully embrace the AI industry for growth.

On 23 June 2025 BusinessGreen published an op-ed by Devorah West exploring how AI’s energy demand could be the final push for the green-energy transition.

Devorah West was invited to Faculty’s AI for Climate Action: Post-COP Reflections and Future Directions event on 4 December 2024 as a panellist, alongside other experts from academia, industry and policy.

An article published on 21 October 2024 on Diginomica discussing sustainable data centres quoted Devorah West on the need for strategic planning and joined-up, cross-departmental thinking on the issue, as well as educating the public on the environmental impact of their own data usage.

Devorah West was featured on BBC Radio 4’s The Artificial Human discussing AI’s energy usage and the importance of making AI green across its whole value chain.

On 18 June 2024 BusinessGreen published an op-ed by Devorah West about the interconnection between the AI and energy revolutions, advocating for governments to stop treating them as separates agendas.

TBI’s report on digital sustainability across the AI value chain (Greening AI) was mentioned on 29 May 2024 on Singapore’s Infocomm Media Development Authority website, CNBC, NBC New York and The National News in articles about Singapore’s new green-data-centre policies. Greening AI was also mentioned the following day in a Tech Bullion article on the same subject.

On 16 May 2024 The New Stateman published an editorial by Ryan Wain and Tone Langengen on the need for new net-zero policies and how AI could be deployed for mass home decarbonisation.

On 28 February 2024, CNA Today published an editorial by Tom Westgarth, Kenddrick Chan and Marie Teo on the strategic importance of Singapore’s National AI Strategy 2.0, which they described as “a pivotal shift in the way Singapore approaches AI”.

Q&A with Devorah West – Senior Policy Advisor, Climate & Energy Policy

How are AI and sustainability related? The AI revolution and the energy transition are deeply intertwined. We need AI to make the breakthroughs needed to fuel the energy transition and we need the energy transition to provide low-cost, reliable energy required to accelerate the AI revolution. Is AI an environmental catastrophe? No. It certainly has an impact on the environment, but by following a green-AI approach we can minimise that impact and ensure that the investments going into AI are used to push forward and accelerate not just AI, but the transition into cleaner, lower-carbon fuel sources and infrastructure. Are you making AI sustainable, or greenwashing it? AI can be sustainable, but we need to work towards a fully “green” AI. The key to doing this is to work across AI’s entire value chain – from the grid to the hardware and software. Specifically, we need to enhance the transparency regarding the energy use and carbon footprint of AI models, in both their training and use. By understanding how much energy is used by different AI models we will be able to hold AI companies and developers accountable – and prevent greenwashing. What is preventing AI from being green already? This issue is twofold: on the one hand, there’s a lack of global measuring standards and, on the other, there’s a lack of support for R&D into more energy-efficient and sustainable AI on an international scale. The international community needs to come together to set up a shared system to measure and report the energy expenditure of AI models and to aggregate resources to support R&D into more energy-efficient AI models. To make sustainable AI a reality, we need people from different disciplines coming together and identifying the solutions to harness both AI and energy transitions on an international scale. That also means decarbonising the grids AI runs on and creating the conditions for investment in clean technologies to scale, as AI companies increasingly operate like energy companies. How do I know if the AI model I’m using is environmentally friendly? At present, we don’t know how much energy is being used by different AI models – and you can’t manage what you don’t measure. We think it’s critical that governments help and incentivise companies to report on the energy usage needed for the development, training and use of AI models.