7 Inclusive Framework for AI Learning at Third Level

Context and Purpose

This Pathfinder research study responds to the growing need for inclusive approaches to artificial intelligence (AI) learning design at third level. As higher education cohorts become increasingly diverse, there is a need to ensure that emerging AI-supported pedagogies are inclusive, accessible, and pedagogically grounded.

Aims and Objectives

• Co-design a novel framework for inclusive AI learning design for neurodiverse students.
• Guide lecturers in the design of inclusive AI instruction strategies.
• Showcase examples of the application of AI in third-level education for neurodiverse learners, contextualised within specific disciplinary domains.
• Maximise the framework’s practicality by providing sample pedagogies and illustrative examples.

Initiative Overview

Song et al. (2024) built a framework for AI learning using the Universal Design Framework guidelines for K-12 level in the USA. Universal Design for Learning is a pedagogical framework that attempts to make learning inclusive for students. The CAST (2018) UDL framework is centered on targeting three key brain networks; the recognition, affective and strategic networks, the What, Why and How of learning respectively. The UDL framework in theory is widely applicable due to its nature, but its basis in K-12 research leaves its potential use outside that environment warranting further research By combining research from all various studies and personal experience working in collaboration with 4^th year psychology students, this Pathfinder case study created a third-level inclusive framework for AI integration in the classroom. Collaborating with 16 academics and 14 psychology students, the co-designed inclusive framework for AI learning, is an extension of the original from Song et al. (2024). It is a proof-of-concept framework under testing to see if it is compatible with third-level educational settings. Some tasks on the outer circle like Assessment of AI-generated content and Review of Said analysis and including AI tools in projects come from personal experience in a classroom. We, as lecturers encouraged our students to use and learn about AI through tasks assigned. Other tasks like identifying problems that AI could assist with learning about models and AI bias and AI human competition are inspired by other research discussed in this case study.

This Pathfinder inclusive framework demonstrates that engagement would be fostered through students embarking on personally relevant projects, choosing topics that resonate with their own experiences and selecting their preferred format of submission (e.g., chatbot dialogue, infographic, or video essay). AI-supported projects or problem-based learning tasks could encourage students to collaboratively analyse phenomena using large-scale datasets processed through AI analytics platforms. In addition, peer and self-evaluation could be enhanced through AI-assisted tools: for instance, feeding a rubric into an AI model to receive constructive feedback, or using AI to rephrase peer reviews with more empathetic, actionable language.

In terms of representation, lecturers could deploy AI-powered tools to present abstract concepts in more engaging, accessible formats. Core content could be transformed into animated visualisations, thereby enhancing clarity for visual learners. AI-powered simulations might allow students to interact with virtual behavioural experiments, manipulating variables and observing outcomes in real-time. AI powered interactive demonstrations could enable learners to explore real-world case studies. To support deeper comprehension, students could access an AI capable of providing layered, personalised explanations and just-in-time assistance when grappling with complex material (Kohnke & Zaugg, 2025).

When it comes to action and expression (the “how” of learning), students could showcase their knowledge using a range of modalities tailored to individual strengths and needs. For instance, those with dyslexia or dysgraphia may lean on tools like Grammarly, ChatGPT or Otter.ai to scaffold their writing and streamline the drafting process. Other students might opt to create artefacts with AI support.

To ensure equity and encourage responsible innovation, authentic assessments could allow students to submit classic deliverables i.e. essays, reports, or case studies, while integrating AI in a clearly documented manner. One marking criterion could assess the transparency and quality of AI engagement, prompting students to critically reflect on how AI contributed to their thinking and output. A student might, for example, use AI to structure their essay outline, summarise readings, or review arguments for clarity, always retaining final editorial control.

Impact and Evidence of Success

Feedback from 16 third level lecturing and support staff involved in teaching and learning has been very positive. They suggested that the framework be extended to cover different subject domains and disciples (art, film etc). 16 psychology students provided qualitative insights into the value of the framework. One student remarked that “overall, AI served as a collaborative partner, useful for support in scaffolding a piece of work and for brainstorming and ideation but it is not a substitute for critical thinking or originality. Another student noticed that “while the AI tool fulfilled the task by identifying the key points, it’s response lacked depth and detailed analysis. This highlights the limitation of relying solely on AI tools ”.

Future Directions and Sustainability

The effectiveness of applying this framework specifically to higher education contexts remains unexplored empirically. While the examples presented illustrate potential, they also highlight  challenges: varying levels of technological proficiency among stakeholders, resource constraints, and potential resistance to pedagogical shifts. This underscores the importance of institutional support, faculty training, and careful alignment between technology integration and curriculum objectives. Future research should empirically validate the framework’s practical impact within diverse higher education settings, clarifying the conditions necessary for its successful implementation, thus refining it to maximise its educational effectiveness.

Alignment and Policy Context

• Institutional Teaching and Learning Strategy
• ESD 2030
• Universal Design for Learning (UDL) Principles
• EU Digital Education Action Plan

Resources and Supporting Information

• Universal Design for Learning Guidelines (CAST)

Images

Funding and Acknowledgements

Funding: SATLE Pathfinder funding