The Role of Generative Artificial Intelligence in Personalised Learning: Opportunities, Challenges, and Future Directions in Higher Education

Abstract

Generative artificial intelligence (GenAI), especially large language models, has enhanced personalised learning by enabling dynamic content creation, conversational assistance, adaptive feedback, and fast resource creation. Nevertheless, technical capability cannot be translated into educational value. This systematic literature review explores the potential of GenAI to improve personalised learning in higher education and determines technological, pedagogical, ethical, and institutional circumstances of implementation. A systematic search of Web of Science and PsycINFO yielded 201 records; following the elimination of duplicates and sequential screening, 18 articles published in 2020-2025 were selected. Thematic synthesis and co-occurrence mapping revealed opportunities in learner-specific content, adaptive tutoring, self-directed learning, accessibility, assessment support, and educator productivity. Constant threats involved hallucinated or biased results, privacy and copyright issues, academic-integrity pressures, unequal access, and cognitive offloading that can undermine critical engagement. There is a lot of evidence that is conceptual, short-term, or grounded on perceived usefulness as opposed to long term learning outcomes. The review supports a human-centred paradigm in which GenAI does not replace but complements teachers. To adopt it responsibly, one needs verified content, redesigning of assessments, open governance, fair access, and clear AI literacy among staff and students. These factors define whether GenAI can offer valuable scaffolding or simply produce answers.