Artificial Intelligence in Education: exploring inclusive use with teachers and educators

This pilot study examines the integration of artificial intelligence (AI) into teacher education as a mediating device for inclusion, rather than as a substitute for peda-gogical mediation. Within the University of Turin’s Expert in Educational and Teaching Processes at School (EsPEDA) Master’s programme (2024–2025), 57 participants (prospective teachers and educators) worked in 12 small groups on two AI-supported laboratory tasks: text simplification and educational storytell-ing. A mixed-methods design combined observations of group dynamics using the Flanders Interaction Analysis Categories (FIAC) with quantitative and quali-tative self-reports. In this sample, FIAC observations indicate a positive collaborative climate. The frequency of exchanges with AI was not associated with satisfaction with the fi-nal product. Conversely, where interaction was more intensive, participants more often reported the need for subsequent human work – source verification, group revision, and adaptation to the classroom context. Text simplification frequently followed a transparent three-stage process (automatic generation → readability check → group revision), whereas storytelling required broader human oversight to address clichés, stylistic uniformity, and conceptual drift. Overall, the findings are consistent with a Universal Design for Learning (UDL) orientation: AI-generated drafts serve as provisional artefacts to be redesigned us-ing professional judgement and appropriate institutional safeguards. Within the limits of this exploratory study, the evidence outlines practical procedures for ini-tial teacher and educator training that balance technological innovation with peda-gogical responsibility.