Digital innovation in EFL through inclusive AI: The University of Macerata case study within the UNITE project

Abstract

Artificial intelligence is transforming multiple domains of education, expanding opportunities for English language learning and teaching while also raising important questions about access, equity, and inclusion (Bahroun et al. 2023). Within this evolving landscape, the Universally Inclusive Technologies to Practice English (UNITE) project investigates how AI dialogue systems can be integrated into English as a Foreign Language (EFL) instruction when their use is guided by the Universal Design for Learning (UDL) framework (CAST 2024). The project aims to understand the pedagogical and technological conditions that enable chatbots to foster engagement, comprehension, and learner agency without introducing additional cognitive or emotional barriers (Fryer et al. 2019; Reinders and Benson 2017; Ushioda 2011; Dörnyei and Ryan, 2015).

This report focuses on the University of Macerata case study, which explores students’ experiences of chatbot-mediated interaction through a structured experimental design. Participants engaged in guided conversations with ChatGPT and Pi.ai, followed by a post-task questionnaire. The analysis concentrates on two key experiential dimensions - immersion in conversational flow and perceived quality of interaction - which closely align with UDL’s three principles of engagement, representation, and action/expression (CAST 2024). Among the project’s main outputs, the Chatbot Accessibility Checklist, developed by the University of Macerata research unit in collaboration with the Center for Applied Special Technology (CAST), the organization that originally designed the UDL framework. The checklist operationalizes these insights into UDL-based criteria for the pedagogical and technical integration of AI chatbots in higher education.