In May 2025 we held two roundtable discussions bringing together college and university staff to explore the intersection of AI and sustainability within tertiary education. These sessions, held in London and Manchester, focused on the challenges and opportunities that AI presents through the lens of the UN’s sustainable development goals, across the themes of environmental, social, and economic sustainability.
The sessions started with a short scene-setting presentation from Jisc. As part of this we look at the UN development goals by theme, including:
Environmental:
- SDG 7: Affordable and Clean Energy
- SDG 13: Climate Action
- SDG 12: Responsible Consumption and Production
Economic
- SDG 9: Industry, Innovation & Infrastructure
- SDG 8: Decent Work and Economic Growth Social
Social
- SDG 4: Quality Education
- SDG 10: Reduced Inequalities
- SDG 5: Gender Equality
- SDG 16: Peace, Justice & Strong Institutions
- SDG 3: Good Health and Well-being
They were then followed by a broad discussion of the three themes, with a change of emphasis on economic sustainably between Manchester and London, switching from broad economics to institutional economics.
Below are the key insights and takeaways that emerged from these conversations.
Our takeaway as that looking at sustainable development through the lens of the UN’s sustainable development goals was useful. We’ll now consider how to build this into our work going forward.
Environmental Sustainability: Striking the Right Balance with AI
AI’s Environmental Footprint:
One of the primary concerns voiced during both sessions was the environmental impact of AI, particularly regarding energy consumption, water usage, and hardware resource intensity required by large language models such as ChatGPT. Many universities in particular have challenging sustainability goals in place. The environmental footprint of AI tools and data centres is significant, and colleges and universities are increasingly looking for ways to reduce this impact while still leveraging the benefits of AI. However, participants pointed out that the discussion around AI and sustainability must be balanced. While AI’s resource usage can be substantial, its ability to contribute positively to environmental goals, such as climate prediction, energy efficiency, and carbon footprint tracking, was also emphasised.
Estimating AI’s Environmental Impact:
One area of discussion was the importance of having data-driven tools to measure the environmental impact of AI. In Manchester, the Hugging Face interactive demo was highlighted as an interesting resource for understanding energy consumption in AI models. However, it’s important to note that many AI tools and models, such as ChatGPT, Copilot, and Gemini, are closed source, with limited information about their environmental impact. While demos such as the one on Hugging Face can help raise awareness, reliable data and transparency from big tech are needed for them to be truly useful.
Mindful Use of AI: Fostering Responsibility and Efficiency
A key theme from both roundtables was the importance of mindful use of AI. It was generally agreed that institutions should encourage students and staff to be aware of AI’s environmental impact and ensure it’s used efficiently.
Participants discussed how institutions should help students and staff reflect on whether AI is truly needed. Instead of using AI out of habit, it’s important to ensure it adds value. Leon Furze’s idea of using local AI on laptops, rather than relying on cloud-based data centres, was mentioned as a way to reduce AI’s environmental footprint, and the counterpoint to this, that data centres are now designed to be as efficient as possible, and shifting AI workloads to local laptops merely shifts the impact, contributing to increases in resource extraction and electronic waste.
By fostering mindful AI use, institutions can balance innovation with environmental sustainability, ensuring AI serves its educational purpose without unnecessary resource waste.
Social Sustainability: Equity, Bias, and AI’s Role in Education
Equity of Access:
In both roundtables, it was pointed out that AI’s rise risks deepening inequalities if institutions don’t take proactive steps to ensure all learners and students can access and benefit from these tools. There was discussion that equal access to AI should be a key consideration as colleges and universities embrace AI technologies in the classroom, and it’s critical to address this issue before access to AI tools become a barrier to some aspects of education. There was also a conversation about digital wellbeing, particularly on how to ensure that AI tools don’t become a source of stress or inequity among students.
Bias and Discrimination in AI:
A recurring theme across both sessions was the inherent bias within AI systems. From image generation tools that perpetuate gender and racial biases to AI recruitment systems that can reinforce discrimination, it was clear that AI must be used with careful consideration of its potential to reinforce societal inequalities. There was a strong consensus that inclusive datasets and more diverse training data are crucial for mitigating these issues. As AI tools become more widespread in recruitment, grading, and assessments, colleges and universities need to establish ethical frameworks to ensure these systems do not exacerbate discrimination. One key takeaway from the roundtables was the need for increased transparency around AI systems’ decision-making processes, especially in assessment. It was also mentioned around recruitment and the need to better prepare students for AI supported recruitment processes.
Ethical AI Use and Academic Integrity:
A strong emphasis was placed on responsible AI use in education, particularly regarding academic integrity. Colleges and universities are still developing clear policies and guidelines to ensure that students use AI ethically and responsibly in their study. There were concerns raised about AI-generated content being presented as original student work, leading to potential misuse in assessments. There was some discussion on how AI could be incorporated into learning in a responsible way, to support students’ intellectual development with the arts sector cited as a good example of mindful AI use:
Digital artists, for instance, often use AI tools to enhance creativity, such as generating visual concepts or assisting in the design process. However, they emphasise using AI as a collaborative tool rather than as a means of simply generating work. The goal is to augment human creativity without replacing it. Artists use AI in a way that is purposeful and intentional, ensuring that the resulting work reflects their artistic vision while still acknowledging the role AI played in the creative process. This approach helps maintain authenticity and artistic integrity while embracing technological innovation.
Impact of AI on Employment:
There was significant concern over AI’s impact on jobs, particularly in the context of displacement vs. transformation. While the fear that AI will eliminate jobs was discussed, it was noted that AI will more likely transform existing jobs—changing the skills required rather than automating jobs. There was discussion around the idea that Colleges and universities should prepare students not only for AI-enhanced roles but also for a workforce that demands a changing skillset. This conversation expanded into labour rights within the AI industry, with a focus on the low-wage workers involved in creating AI systems, and how institutions can advocate for better working conditions for these workers. This ties into broader issues of human rights and decent work within the AI supply chain.
Economic Sustainability: AI as a Tool for Institutional Efficiency
AI’s Role in Enhancing Efficiency:
Both roundtables highlighted the potential of AI to drive institutional efficiency. For example, AI-powered automation tools like Teachermatic (for marking) were discussed as cost-effective solutions that can significantly reduce administrative workload and improve staff wellbeing. By automating repetitive tasks, AI frees up time for educators and administrators to focus on more critical activities, improving the overall efficiency of institutions, for example on participant shared how Freshworks can handle student queries and improve operational efficiency. Additionally, the potential for AI to contribute to energy efficiency through predictive maintenance and smart campus management was highlighted. AI tools could optimise energy usage, resource allocation, and campus operations, which not only helps reduce costs but also supports the universities’ sustainability goals.
There was also a call to consider how AI could be integrated with existing digital transformation projects to enhance overall institutional productivity, particularly in areas like research support and student services.
AI Procurement and Vendor Engagement:
A major concern raised at the Manchester roundtable was the disconnect between IT departments, learning technologists, and sustainability teams when it comes to AI procurement. It was noted that AI procurement is often happening in isolation from sustainability goals. Colleges and universities need to integrate sustainability into their AI procurement frameworks to ensure the tools they adopt are aligned with their environmental objectives.
One suggestion was that as metrics aren’t readily available, institutions should ask open-ended questions that encourage transparency, such as how aware suppliers are of their environmental impact and how they plan to share this information. This helps establish a dialogue on sustainability, even when exact metrics are not yet available.
The Crescent Purchasing Consortium was mentioned as a specialist sustainable procurement training provider, alongside Ethical consumer in Manchester who work with people to integrate sustainability into their purchasing decisions and Net positive futures initiatives, who help with carbon reduction plans.
There was also a call to engage with smaller AI vendors who might be more flexible and socially responsible in terms of sustainability compared to larger corporations. Institutions are encouraged to ask open-ended questions during procurement about resource usage, carbon impact, and supplier transparency.
Sector-Wide Collaboration: A Unified Approach
One of the major themes from both roundtables was the need for sector-wide collaboration to tackle the complex issues surrounding AI and sustainability. Colleges and universities can amplify their impact by sharing best practices, lobbying for better policies, and working together to create coherent AI frameworks that balance sustainability and educational integrity.
Institutions could also come together to demand more transparency from AI suppliers, especially about their environmental and social practices. By collaborating and working as a sector, colleges and universities have more power to drive change across the AI industry and ensure that sustainability remains a central part of AI adoption.
Actions Identified for Jisc
- Continue to provide guidance on the responsible use of AI, helping institutions implement best practices for sustainable AI adoption.
- Identify the right questions for universities to ask suppliers about the environmental impact of AI tools, particularly around energy consumption and resource use.
- Facilitate sector-wide conversations and collaboration, enabling institutions to share best practices and collaborate on developing a unified approach that embraces AI and aligns with their sustainability goals.
Conclusion
The idea of broadening the discussion on sustainability and look at through the lens of UN’s sustainable development goals, was from our perspective, fruitful and construction. We’re aware that the conversations around AI and sustainability almost certainly centre on environmental sustainability, and it’s fair to say, from the open statements from participants that this was the expectations of this session. We thank the participants for joining on a broad and very enjoyable couple of session.
As AI continues to evolve and shape the future of tertiary education, institutions have to find a way to balance the innovative potential of AI with their commitment to environmental sustainability. One of our aims of the session was to get ideas on how we could frame our work to ensure a balanced overall approach to sustainable development, and we were pleased that roundtable participants contributed to this broad discussion.
The discussions from the London and Manchester roundtables highlighted the need for colleges and universities to approach AI responsibly, focusing on both its positive contributions and its challenges. By fostering collaboration, promoting ethical AI use, and providing clear guidance on sustainability, institutions can unlock the full potential of AI across the full range of the UN’s sustainable development goals.
AI Tools Mentioned:
- Generative AI (e.g., ChatGPT): LLMs discussed for their energy consumption and environmental impact.
- Hugging Face ChatUI-energy: A resource for estimating the energy consumption of AI models.
- Teachermatic: For automating grading and feedback, improving efficiency and staff wellbeing.
- Freshworks: Used for handling student queries and improving operational efficiency.
- Microsoft Copilot: For improving productivity in university workflows.
With thanks to our roundtable participants:
- Peter Cuthill Royal Northern College of Music
- Sophie Hughes-Saunier Coleg Cambria
- Sara Kassam King’s College London
- Anna Klemming University of Westminster
- Tim Leonard Lancaster University
- Dominic Pates City St George’s, University of London
- John Ray University of Oxford
- Chris Rowell University of the Arts London
- Amanda Seys Harper Adams University
- Monica Westin Manchester Metropolitan University
- Sharon Whitehouse Westminster Adult Education
- Donna Wilson Coventry College
Find out more by visiting our Artificial Intelligence page to view publications and resources, join us for events and discover what AI has to offer through our range of interactive online demos.
Join our AI in Education communities to stay up to date and engage with other members.
Get in touch with the team directly at AI@jisc.ac.uk