ID128 Building Institutional AI Capability through an AI Acceleration Centre

The Idea

Despite the transformative potential of AI, its use by students’ remains largely confined to low‑level functions such as explaining concepts or summarising text suggesting little impact on higher‑order learning (HEPI 2025). To achieve genuine transformation, AI must be purposefully integrated into curricula. However, this requires customising AI technology to ensure responses and information generated by the tools are specific to the course and students’ needs. Leading UK universities have already started this process: the University of Manchester’s NebulaONE platform allows educators to create and customise AI models in a secure environment and without needing technical skills.  At Keele, emerging evidence from faculty-level surveys and workshop feedback suggests that staff are interested in using AI to support genuine pedagogical innovation, rather than simply as a productivity tool. For example, colleagues in the Faculty of Medicine and Health Sciences are exploring ways to develop GenAI-enabled simulated patients and authentic educational videos to enhance teaching and learning. However, these more advanced applications require access to multimodal AI models with customisation capabilities, features that are not currently available through Copilot. Institutionally, there is also no clear mechanism to support staff in these more advanced uses of AI. Together, these limitations constrain the University’s ability to explore more ambitious, discipline-specific applications of AI and highlight the need for a supported institutional environment in which such innovations can be developed and evaluated. To strengthen the evidence base for the idea, a needs analysis will be conducted to identify current staff capability, areas of interest, and priority support needs in relation to more advanced uses of AI. While addressing this will ultimately require institutional investment, it would be risky to commit at scale without a clear evidence base. Instead, a phased, evidence-generating approach is needed to inform strategic decision-making. This would also support a more ethical and sustainable model of adoption. Rather than promoting widespread access to advanced AI tools that may not be needed or used by all staff, this approach would target access to those areas where there is clear pedagogical value and a defined need. Once established, consideration could also be given to evaluating whether the use of advanced AI offers a more efficient and sustainable alternative to existing practices, for example where it may reduce reliance on more resource-intensive manual processes 

We propose that Keele establish an AI Acceleration Centre, initially as a pilot-phase initiative led by a small specialist AI and Innovation Team. In the first instance, this team could be formed through a targeted combination of existing colleagues supported through partial secondment and collaboration with external partners such as the Institute of Technology, rather than requiring a wholly new staffing structure at the outset. This would help keep initial costs relatively modest, with expenditure focused primarily on a limited number of advanced AI subscriptions and staff time, rather than on establishing a large new unit from the outset. The centre would provide access to a selected suite of advanced AI technologies within a supported environment in which the team and early adopters could develop and test innovative applications. The team would support colleagues, particularly those involved in curriculum development or those wishing to explore AI as part of their scholarship, to identify ideas that are pedagogically relevant and feasible within existing programme or module structures. In this way, development and piloting could be aligned, where possible, with ongoing curriculum enhancement, scholarship activity, or planned module review, rather than relying entirely on additional staff capacity. 

Why This Idea Should Be Considered

This proposal is worthy of consideration because it presents a vision for transformative AI adoption for the University, while offering a practical, low-risk, and cost-effective pathway for implementation. It supports the University’s commitment to promoting digital capability and will help form the university’s leadership for advancing AI practices. It is needed to cultivate the ideas and visions of the early adopters within the University while simultaneously generating the evidence needed to guide future investment decisions. 

How We Would Implement This Idea

To implement the idea: 

1. Establish a multidisciplinary AI and Innovation Team to lead the Centre and provide expert support. This could draw on existing colleagues with relevant experience in education and learning technology, with additional expertise and input from the Digital Society Institute and external partners such as the Institute of Technology. 
2.  To begin with, secure paid subscriptions to a small suite of advanced external AI models for pilot use, rather than developing a Keele-owned model at this stage. This would provide a practical, lower-risk route to piloting innovation while avoiding the cost and complexity of building an institutional model too early. Current tools are typically available on a monthly subscription basis, at approximately £18–30 per month for a small team. The initial focus would be on models that support multimodal working, safe customisation, and discipline-specific application. 
3.  Create a structured access and support system enabling interested colleagues to book consultations, explore ideas, and test selected models in a supported environment. A dedicated AI suite, equipped with appropriate PCs and located in a visible and accessible setting, would make this approach practical and scalable. There may also be value in linking aspects of this activity to Keele in Town, providing opportunities for wider civic impact by helping the local community learn about AI and how it can be used ethically in everyday life.

What Success Would Look Like

Initial indicators of success will focus on educational impact, since it is anticipated that the majority of the applications developed will be education related. 

1. Although policies on AI use are increasingly in place, there remains limited practical guidance across the sector on how AI can be integrated across programmes in ways that support the progressive development of students’ AI capabilities relevant to both academic performance and employability. The Centre would help address this gap by developing and evaluating practical approaches with clear institutional value. For example, this could include programme-level frameworks for introducing AI capabilities progressively across stages of study as well as practical guidance to support staff in developing and advancing their own AI capabilities. In doing so, the Centre would aim to generate evaluated examples of practice that support responsible and effective use of AI, while also providing a clearer basis for future institutional decision-making. 
2. Demonstrable positive impact on student learning experience and academic performance, evidenced through evaluation data and assessment outcomes. 
3. Generation of high-impact case studies for publication and dissemination, establishing the AI Acceleration Centre as a strategic institutional hub for educational transformation.