AI-Enhanced Audiobook Adoption Among Differently-Abled University Students in Sri Lanka: A Systematic Review and Theoretical Integration

The rapid integration of Artificial Intelligence (AI) into educational technologies has significantly expanded opportunities for inclusive learning. Among these innovations, AI-enhanced audiobook systems featuring neural text-to-speech synthesis, adaptive narration, personalized content recommendations, and intelligent summarization offer transformative potential for differently-abled university students. Despite global advancements, limited empirical and theoretical synthesis exists regarding audiobook adoption in developing-country higher education contexts, particularly in Sri Lanka. This paper systematically reviews interdisciplinary literature spanning digital accessibility, assistive technologies, audiobook pedagogy, technology acceptance theories (TAM and UTAUT2), disability studies frameworks, and Structural Equation Modeling (SEM) applications. The review follows PRISMA 2020 guidelines and is grounded in a documented search protocol across five electronic databases. The review identifies critical theoretical gaps, including the absence of integrated AI accessibility adoption models and limited South Asian empirical validation. Building upon these findings, the paper proposes an extended UTAUT2-based conceptual framework incorporating AI-generated personalization and Institutional Readiness as novel constructs. The proposed framework is visualized as a structural path model and discussed in relation to its implications for disability services, edtech design, and inclusive education policy in Sri Lanka and comparable developing country contexts. The study contributes to technology adoption scholarship, inclusive education policy discourse, and AI-driven accessibility research.