A UDL–ICF Framework for Inclusive Artificial Intelligence in Cultural Heritage: Exploratory Insights from an Erasmus+ Blended Intensive Programme

Cultural heritage institutions continue to face significant challenges in addressing neurodivergent access needs, while a principled framework linking inclusive design and artificial intelligence in heritage remains underdeveloped. This article proposes and explores the transfer of a dual Universal Design for Learning (UDL 3.0)–International Classification of Functioning, Disability and Health (ICF) framework from education to cultural heritage. The framework was examined within an Erasmus+ Blended Intensive Programme in Naples, Italy, through an exploratory mixed-methods pilot design combining pre/post questionnaires with thematic analysis of participants’ open-ended reflections. Findings suggest that the programme supported a qualitative reframing of accessibility from a predominantly physical-access model toward a broader understanding encompassing sensory, cognitive, communicative, and participatory dimensions. Site visits to Pompeii, Herculaneum, and the MANN emerged as especially important in prompting critical reflection on the gap between heritage value and accessibility practice. Quantitative patterns also indicated preliminary gains in participants’ self-efficacy in applying UDL/ICF principles and in reflecting on AI ethics in heritage, although these findings remain exploratory given the small sample size and pilot design. The article contributes a conceptual and pedagogical framework for inclusive AI in cultural heritage and discusses its implications for accessibility governance, conservation-constrained settings, and future co-designed research.