Decoupling Personalised and Adaptive Learning in AI-Enhanced Education: A Narrative Review and Conceptual Clarification

Abstract

Personalised learning and adaptive learning are often treated as the same idea in AI-enhanced education, even though they represent different pedagogical purposes. This narrative review examines ten contemporary studies, including work on intelligent tutoring systems, adaptive platforms, multi-agent AI, and human-centred AI frameworks, in order to clarify the distinction between these two concepts. The review shows that personalised learning involves intentional, human-centred customisation based on learner goals, preferences, identity, and context. In contrast, adaptive learning relies on algorithmic, real-time adjustments based on performance data. While AI systems are increasingly sophisticated in adaptivity, they rarely achieve genuine personalisation, often limiting learner agency and overlooking socio-emotional or cultural dimensions. This review also introduces a conceptual model that positions personalisation and adaptivity as related but distinct constructs requiring different pedagogical intentions, data models, and levels of human oversight. Understanding this distinction is essential to prevent misinterpretation, guide responsible AI adoption, and ensure that automation supports rather than replaces human judgement. The paper concludes with recommendations for educators, instructional designers, and AI developers to integrate AI in ways that are transparent, ethical, and centred on meaningful learner autonomy.