Access for All The promise and perils of assistive AI

Access for All? The promise and perils of assistive AI Wednesday 27 May 2026 AI is redefining accessibility, but who is shaping it, and how can inclusion be built in from the start? AI-assisted technologies are reshaping how people with disabilities access information, navigate spaces and participate fully in education, work and society. As these technologies have become more widespread, important questions have emerged about who they serve, how they are designed, and whether they genuinely deliver inclusive access. This event brought together experts from technology, advocacy and lived experience to explore the promise, and the limits, of AI-assisted technologies. As AI has opened new possibilities for personalised support, alternative modes of interaction and wider participation in work, education and culture, it has also raised critical questions about bias, agency, privacy and responsibility. Through informed discussion, the panel examined how AI can genuinely improve accessibility without reinforcing exclusion or shifting burdens onto individuals, and what it will take, technically, ethically and institutionally, to ensure these tools are designed with, not just for, the people they are intended to support. Speakers Introduced by: Professor Susan Downes, Professor of Ophthalmology, Nuffield Department of Clinical Neurosciences Moderated by: Ben Tillyer, Head of Digital Accessibility, University of Oxford Panellists: • Hector Minto, CEO of Kerbcut; former Director of Commercial Accessibility at Microsoft • Charlie Turrell, Accessibility Champions Manager and Engagement Lead at the Ministry of Justice; founder of CAN and Director of Ab11y.com • Dave Williams, Chair of the Board of the Braillists Foundation; Customer Experience Manager at Dot Inc; former Inclusive Design Ambassador at RNIB