Universal Design for Learning (UDL) offers a valuable framework for developing inclusive and accessible courses, but educators often receive limited support in translating its broad principles into context-specific teaching and assessment practices. This paper reports on the design of a custom generative artificial intelligence (GenAI) conversational assistant prototype developed to support computing educators in Australian tertiary education in applying UDL to their courses. The prototype provides concrete, non-directive suggestions across the three tenets of UDL, multiple means of engagement, representation, action and expression. Rather than treating GenAI as a generic automation tool, the prototype explores how prompt design, scope limitation, structured outputs, and persona-informed contexts can scaffold inclusive pedagogical reflection while maintaining academic judgement. The paper presents the functional prototype as its primary contribution and describes its design considerations, prompt architecture, and early scenario-based testing.
Keywords
Universal Design for Learning (UDL), generative AI, inclusive education, computer science education, course design, Australian Universities, Design-based research.
Author Biographies
Dr. Anne Ozdowska, The Australian National University
Dr Anne Ozdowska is a Research Fellow in the School of Computing at the Australian National University. Her research explores how interactive technologies can take on companion-like roles and how these relationships are shaped through physical presence, sensory interaction, responsiveness, and everyday routines. She is particularly interested in the design of embodied and tangible AI companions that support comfort, connection, care, and meaningful engagement without displacing human relationships or reducing companionship to conversation alone. Her work examines how people interpret the agency, intentions, and behaviours of companion technologies, as well as the ethical tensions that arise around attachment, dependence, trust, and responsibility. Anne also uses participatory and design-led methods to investigate how AI companions might be designed in ways that are inclusive, context-sensitive, and attentive to different forms of social and emotional need.
Professor Penny Sweetser Kyburz, The Australian National University
I lead the ANU Games Research Group and teach Game Development. I am a researcher in the areas of AI, player experience, and human-AI interaction in games and mixed reality and an experienced game developer. I developed the GameFlow model, which is a widely cited model of enjoyment in video games, and authored a book on Emergence in Games. I have contributed to 20 digital games, which have garnered many awards, including four prestigious BAFTA nominations. I have devised usability methods and lead usability testing on AAA games, resulting in very high critic scores. I have also worked in the Senate as a policy adviser in digital rights, communications, and disability. I am an advocate for diversity and inclusion in technology and games.