In early June, we hosted two roundtables, one in London and one in Manchester, bringing together voices from across colleges and universities to explore the UK Government’s AI Opportunities Action Plan.
These events offered rich insights into how institutions are already active in some of the areas covered by the plan, and how participants viewed some of the recommendations. While the tone varied, with London tending toward critical reflection and Manchester leaning into practical innovation, both sessions surfaced shared ambitions, probing questions and a strong sense of sector responsibility. What emerged was a mix of cautious optimism, practical experience, and calls for change.
Our sessions began with a short introduction from Michael Webb, Director of AI at Jisc. He took a moment to put the discussions into context, sharing an overview of the UK Government’s AI Opportunities Action Plan and what it might mean for colleges and universities.
This plan is being positioned as a catalyst for economic growth, skills development, and public service transformation. It focuses on three big priorities: laying solid foundations (things like data, skills, infrastructure, and trust), helping people see the benefits of AI in their everyday lives, and building a strong, homegrown AI future for the UK.
Michael highlighted some of the most relevant commitments in the plan for the tertiary education sector. These included an ambition to boost the UK’s computing capacity twentyfold, develop new AI Growth Zones, and open up high-value public datasets to drive innovation and research. When it comes to skills, the plan is aiming high: it wants to see more AI professionals trained, better support for universities to offer industry-relevant courses, new pathways through further education and apprenticeships, and meaningful action on diversity and inclusion.
Framing the Discussion
To help kick-start the conversation, Michael posed a few questions for everyone to reflect on:
- How can the sector play its part in creating a more diverse and inclusive AI talent pipeline?
- What role should colleges and universities take in shaping and making the most of national data assets?
- How should institutions respond to the UK’s infrastructure ambitions, like the growth of AI zones and international computing partnerships?
These questions helped participants think about what is already working well, where the gaps might be, and what more the sector could do to help shape the UK’s AI landscape in the years ahead.
Here are the key takeaways, that emerged from these conversations organised into four themes.
Skills and Talent: Redefining What It Means to Be AI-Ready
Both roundtables reinforced that the UK is at a crossroads when it comes to AI-related skills. There was a widespread view that the Action Plan leans too heavily towards technical AI developer roles, and not enough on the broader skills needed to use AI effectively.
In London, participants questioned what AI skills actually are. Is it coding models, or is it about understanding AI’s implications for employment, ethics, and everyday work? There was a strong push to broaden the definition to include critical thinking, data literacy, ethical awareness, and practical application. Staff and students alike need to understand how to use AI well, not just how it works.
There was some discussion around the idea that AI-related skills don’t sit neatly within traditional subject boundaries. While technical skills remain important, the real gap may be in people who can critically evaluate, apply and communicate AI, and support others to do the same. Educators themselves need to become ‘AI translators’ and they’re not being given the time, training or support to do so.
Several contributors noted that FE colleges are often leading the way in delivering practical AI training, not just for learners but also for local businesses. One college, for instance, rolled out a 3-hour GenAI course for SMEs and built it into wider upskilling programmes for staff. Lifelong learning was a common thread, especially in Manchester’s session, where FE’s role in building digital resilience for communities was highlighted.
Another key challenge is the lack of confidence among educators, especially where students have become more AI savvy than staff. Some are already using AI to generate work, even where assessment policies are unclear. But as one attendee noted, ‘assessment change takes a panel and a process’ demonstrating the distinct lack of sector agility.
There is deep concern about the digital divide, some learners arrive with a working knowledge of image generation tools, while others don’t even have a smartphone. As one participant said, ‘How do we teach to both?’
Next Steps for the Sector:
- Reframe AI skills beyond technical roles.
- Scale up staff development, particularly in ethics, critical thinking, and use of AI tools.
- Partner with employers to align AI training with real-world roles and needs.
Data: From Collection to Meaning
Across both roundtables, data emerged as areas of high ambition but low clarity. In London, there was strong unease about how data is currently collected and used, particularly in a context where students and staff routinely tick consent boxes without understanding the implications. Many felt the Action Plan’s emphasis on public datasets must be underpinned by robust ethical frameworks, opt-out mechanisms, and clear inclusion of student voice to avoid further eroding trust.
Participants pushed for a more human-centred and meaningful approach to data. One suggestion, relating to data use in education, involved students writing a letter to their future selves during induction, capturing their motivation and aspirations. When combined with engagement metrics, feedback loops and wraparound support, such qualitative data could offer richer insight than traditional performance measures. Others highlighted the value of understanding what learners enjoy, how they prefer to engage, and what brings them joy, arguing that this kind of insight is often overlooked but critical to meaningful education outcomes.
In Manchester, there was greater appetite for a national lifelong learning dataset, but only with guardrails in place. Attendees raised sharp questions: who would own this data, and who would benefit? Could it be at risk of surveillance or profiling, or could it be a powerful tool for personalised support and fairer outcomes? There was strong support for making insights more accessible, for instance through conversational interfaces, chatbots that could query datasets without the need for complex dashboards or coding skills.
However, many emphasised that data alone isn’t the solution, it’s the questions we ask of it and the judgements we make based on it. There was concern that predictive data use such as identifying students at risk of dropping out could narrow opportunity, whereas supportive data use such as tailoring support based on student motivation might open new doors. Data systems should empower, not sift.
Several contributors warned of a significant data literacy gap across staff, students and the public. Without the skills to question, interpret and challenge data, even well-intentioned systems could reinforce inequality. There was interest in tools that made data reflective and personal, such as a ‘Spotify Wrapped’ style learning reflection, though participants noted these would need watertight ethical governance.
Libraries were highlighted as critical but underused partners, not only as data providers, but as intermediaries who could help navigate licensing, ethics, and system design.
Next Steps for the Sector:
- Build ethical, learner-centred frameworks for data use and transparency.
- Work with libraries, communities and learners to improve data literacy and agency.
Infrastructure: Scaling with Purpose
Participants questioned the government’s ambition to scale compute power twentyfold, both from a sceptical viewpoint – given the energy, cost and environmental implications, but also whether it was actually ambitious enough. There was strong interest in aligning AI Growth Zones with green energy and local regeneration, and in learning from live examples such as Milton Keynes’ bid to become an AI innovation hub. There was, however, scepticism around the idea that co-location of data centres and AI companies was actually useful.
There was broad agreement that while the UK cannot and should not try to out-compete the likes of OpenAI or Google, it could lead in building ethical, locally relevant AI trained on UK-held data. In this context, local specialism, public trust, and institutional partnerships were seen as genuine competitive advantages.
Finally, participants urged caution: AI infrastructure decisions taken now could shape not just education but employment, equity and public services for decades to come. If the sector is to engage meaningfully with the government’s infrastructure ambitions, it must be at the table early, often, and with a strong voice.
Next Steps for the Sector:
- Collaborate with local partners to support bids for AI Growth Zones.
- Advocate for sustainable, ethical infrastructure that benefits communities.
Changing Lives: Structural Change or Surface-Level Action?
There was a view that changing lives with AI would only work if diversity, accessibility and inclusion were considered. Participants were wary of performative gestures and emphasised the need for systemic reform. Structural barriers, not just lack of interest, are excluding diverse groups from AI careers
Discussions underscored that inclusion isn’t about fixing individuals, it’s about rethinking systems. Participants called for a rethink of how we brand, teach, and talk about AI careers. They argued for AI as a tool for everyone, not just a field for the technically elite. If AI is to be truly transformative, its development must include those most impacted by automation, bias, and exclusion.
Participants highlighted how the dominant culture in tech, rooted in patents and the myth of the ‘lone genius’, can feel exclusionary and heavily male dominated. They also pointed to the failure of education systems to support diverse learners early enough in their journeys, by the time individuals reach further or higher education, many have already been excluded.
There was recognition that bias in AI systems stems from societal bias, and tackling it requires more than technical fixes, it demands cultural and structural change.
The Manchester group picked up on the branding issue, that AI and computing are still seen as ‘techy’ or ‘nerdy’. Some advocated to rebrand courses such as ‘Digital Creative’ and change them to ‘Product Design’ to help improve the gender . Others noted how AI opens the door for new, more accessible pathways into digital careers.
Next Steps for the Sector:
- Redesign careers outreach and course branding to reflect AI’s broad applications.
- Fund diverse PhD pipelines and ensure ethical oversight for AI research.
- Include staff, students and communities in shaping what ‘Responsible AI’ looks like in tertiary education.
Actions identified for Jisc:
- Continue to provide national coordination for ethical, inclusive AI adoption
- Support staff AI skills development at scale
Final Reflections
Across both roundtables, one message came through loud and clear, the sector is already moving, but it needs joined-up national leadership, grounded in reality. The tertiary sector isn’t waiting. Colleges and universities are already innovating, with limited resources and impressing creativity. Jisc could provide that national coordination and help translate policy ambition into practical action, facilitating collaboration across the tertiary education sector and amplifying the sectors voice.
Many thanks to participants:
Associated Community Training Limited (ACT): Sam Holland
Coleg Cambria: Nigel Holloway
Coventry College: Judy Kay
Hertford Regional College: Olive Oliver
Milton Keynes College: Penny Langford
Nelson and Colne College: Fionnuala Swann
South Staffordshire College: Steve Wileman
University College London: Jason McEwen
University of Leeds: Samantha Pugh
University of London: Richard Michel
University of London (Worldwide): Christina Koziol Webster
University of Oxford: Amy Warner May
University of Reading: Melanie Stockton-Brown
University of the Arts London: Christie Johnson
University of the Arts London: Ruth Powell
University of York: Steve King
York College & University Centre: Abby Parkin
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
Was this right? It doesn’t it to me