AI ‐Mediated Multimodal Reading Education for EFL Learners: Effects on Students' Enjoyment and Engagement

ABSTRACT Artificial intelligence (AI) has increasingly reshaped educational practices by expanding opportunities for adaptive, multimodal and learner‐centered instruction. While growing attention has been paid to AI‐supported language learning, less is known about how multimodal AI technologies may support the emotional and behavioural dimensions of education among at‐risk students, particularly in second/foreign language reading contexts. Addressing this gap, the present study examined whether AI‐mediated multimodal reading education could enhance at‐risk English as a foreign language (EFL) learners' enjoyment and engagement during reading instruction. Using an interventional pre‐test/post‐test design, 80 at‐risk EFL students were assigned to either a control group ( n = 40) or an experimental group ( n = 40). Over two semesters, students in the experimental group received reading instruction supported by multimodal AI reading bots and AI chatbots, whereas those in the control group received conventional reading instruction. Enjoyment and engagement questionnaires were administered at the beginning and end of the intervention. ANCOVA results showed that students in the AI‐mediated education group demonstrated significantly higher levels of enjoyment and engagement than their counterparts in the control group. These findings suggest that multimodal AI technologies can contribute to more inclusive and emotionally supportive reading education for at‐risk learners. The study offers implications for educational researchers, teachers and policymakers seeking to integrate AI technologies into equitable and engaging language education.