Teaching Literature with AI: A PRISMA-Aligned Review of Pedagogically Mediated AI Approaches in EFL University Classrooms

Abstract

The rapid development of AI has generated interest in its application to higher education, including the teaching of literature in EFL. However, research on AI integration in university-level EFL literature education remains fragmented and under-theorised, particularly regarding pedagogical and ethical considerations. This study presents a PRISMA-aligned systematic review of peer-reviewed research examining AI-driven approaches in EFL literature classrooms. Analysis was guided by an integrated conceptual framework incorporating sociocultural theory, reader-response theory, constructivist learning theory, and human-centred AI principles. The review identifies AI tools—including natural language processing applications, generative AI systems, conversational agents, and adaptive learning platforms—as most effective when pedagogically mediated and ethically governed. Key educational outcomes include enhanced literary engagement, development of interpretative skills, increased learner autonomy, and emerging critical AI literacy. Challenges involve maintaining interpretative authority, addressing algorithmic bias, ensuring academic integrity, and providing institutional support. Overall, AI can augment EFL literature education when positioned as a mediational resource within theoretically grounded, ethically responsible instructional designs. The findings offer implications for curriculum development, teacher education, and policy, while highlighting the need for longitudinal, theory-driven, and context-sensitive research.