Multimodal Reading Engagement: A Conceptual Review of AI-Mediated Text Transformation

Abstract

The increasing availability of generative artificial intelligence (AI) tools has introduced new possibilities for transforming how students interact with academic texts. While most educational discussions have focused on AI-assisted writing and assessment, the role of AI in supporting reading engagement remains underexplored. This study presents a narrative literature review examining the relationship between generative AI, multimodal learning, and reading engagement in higher education. Relevant studies were identified through academic databases using keywords related to generative AI in education, multimodal learning, and reading engagement, and were analyzed through thematic categorization. The review indicates that AI-mediated text transformation, particularly converting text into visual or animated representation, may restructure reading activity into an iterative process involving rereading, verification, and adjustment. Such interaction aligns with cognitive, behavioral, and affective dimensions of engagement by encouraging learners to compare generated representations with original textual sources. Rather than functioning as an automated comprehension tool, generative AI can act as a mediating learning artefact that supports active involvement with reading material when pedagogically guided. The study proposes a conceptual framework describing AI-assisted text-to-animation learning as a cyclical engagement process. The findings suggest that the educational value of generative AI depends on instructional design that promotes critical interaction with generated output. Future research should empirically investigate classroom implementation to validate the proposed framework.