Is “Sustainable AI” a contradiction in terms? This was the question at the heart of our first “AI in Motion” event, a series tackling big human questions around AI. The session began with a debate chaired by TLT’s Tom Sharpe, with contributions from experts from Natasha McCarthy (Royal Academy of Engineering) and Robert Keus (GreenPT), followed by a panel of expert speakers in Emily Farrimond (Baringa Partners), William Quan (Fleete Group Limited) and Andrew Burgess (Greenhouse Intelligence), chaired by Emma Erskine-Fox of TLT.
The consensus: Sustainable AI is not only possible, but essential. Here’s how organizations can bridge the gap between ambitious AI goals and net zero commitments.
The scale of the challenge: be realistic
The rapid proliferation of AI poses undeniable challenges for decarbonization. The energy and water requirements of AI models are significant and reliable measures of sustainability are still difficult to find. As Andrew Burgess noted, “One of the biggest challenges is the lack of information and measurements. We’ve talked about some of the metrics available to people, but does anyone really believe in these metrics? The people who generate them are also the people who sell the models. So it’s difficult to balance until there is a lot more transparency about the sustainability data available.”
Without reliable data, it is difficult to make meaningful comparisons and make informed decisions.
To remember:
Recognize the scale and complexity of aligning AI adoption with net zero ambitions. Honest conversations about challenges are the first step toward solutions.
Creative solutions: be creative
Sustainability in AI isn’t just about technology: it’s also about mindset. The panel highlighted the need to be creative in how we develop, procure and integrate AI. This means exploring smaller, more efficient models, asking tough questions of suppliers, and integrating sustainability into policy, governance, and training. As Emma Erskine-Fox says: “Organizations can look at many ways to reduce the environmental impact of their use of AI by being creative in how they develop AI solutions, in how they purchase them, by integrating sustainability into training, policy, guidance and governance frameworks around the use of AI internally. »
To remember:
Think outside the box: Innovative approaches to AI development and deployment can significantly reduce environmental impact.
Transparency and accountability: being responsible
Transparency is a recurring theme. Developers need to be honest about the resource and energy consumption of AI systems. Sustainability considerations span the entire AI value chain, from hardware and infrastructure to data centers and use cases. Responsibility is shared: developers, organizations and end users all have a role to play. As Robert Keus said: “As human beings, as unique users of AI, we have our own responsibility for how we should use it. » And Natasha McCarthy added: “This shows how the types of responsibilities and decisions are distributed to make AI and its use more sustainable. »
To remember:
Sustainable AI requires end-to-end accountability and greater transparency at every stage, from development to individual use.
Empowering Users: Be Informed
Accelerating the responsible adoption of AI means empowering users. Training is essential, not only to know how to boost AI, but also to understand its true costs and capabilities. This builds trust and ensures that AI is used where it adds real value. William Quan highlighted, “How do we accelerate adoption in the right way, but also how do we train end users to be able to invite, but also understand the technology in a way that maximizes efficiency and productivity?
To remember:
Invest in user training to maximize the benefits of AI and minimize unnecessary energy consumption.
Hope for the future: stay hopeful
Continued dialogue, improved data quality and progress towards common standards for emissions reporting offer real reasons for optimism. As Tom Sharpe pointed out, these developments will enable “adopting the right technology in the right way”.
Despite the challenges, there are real reasons for hope. As Emma Erskine-Fox summed it up so well: “Be realistic, be creative, be hopeful. »
To remember:
Progress is possible: by working together, we can harness the potential of AI for beneficial purposes while preserving our planet.
Final Thoughts
Sustainable AI is not a contradiction. It is a collective challenge that requires realism, creativity, transparency and hope. By adopting these principles, organizations can ensure their AI ambitions support, rather than undermine, their sustainability goals.
Key points to remember:
- Be realistic about the scale of the challenge.
- Be creative in finding solutions.
- Be transparent and accountable at every step.
- Empower users through education.
- Stay hopeful: progress is within your reach.
This publication is intended to provide general guidance and represents our understanding of the law and practice relevant to October 2025. Specific advice should be sought for specific cases. For more information, see our terms and conditions.