AI-Assisted Audio Learning In Educational Settings: Opportunities, Challenges, And Future Research Directions

Abstract

This systematic literature review examines how audio-based artificial intelligence (AI) tools are being used to support learning across educational contexts and what challenges and opportunities arise from their implementation. Using a structured search across major e-scientific databases, this study identified and analysed empirical and conceptual works on AI assisted audio learning, including audio feedback, voice tutors/assistants, adaptive audio modules, and AI generated podcasts. VOSviewer was employed to map keyword co-occurrences and thematic clusters, while a large language model (LLM)-based notebook was used to assist in summarising abstracts, methods, and key findings, and to group studies by educational level and AI technology type. The findings indicate that audio-based AI interventions generally contribute positively to cognitive outcomes, motivation, and engagement, particularly by enabling frequent, personalised feedback and flexible, on-the-go learning opportunities. At the same time, the review reveals several challenges related to instructional design, teacher orchestration, data privacy, and the risk of learner overreliance on AI feedback. The existing evidence is heavily concentrated in language learning and a few specialised domains, with a predominance of short-term studies and self-report measures, limiting the generalisability of current conclusions. This review contributes by offering an updated mapping of AI-assisted audio learning research, highlighting underexplored contexts and design configurations, and outlining clear directions for future studies, including longitudinal designs, comparative trials of audio versus text-based feedback, and investigations into how teachers and students negotiate the role of AI in classroom practice.