Generative AI-enhanced adaptive learning for improving student learning outcomes: a systematic literature review

Abstract

This study systematically reviews the use of generative artificial intelligence (GenAI) in adaptive learning to support measurable student learning outcomes. The revised focus responds to methodological concerns by operationalizing the outcome variable as learning outcomes, including conceptual understanding, critical thinking, digital literacy, problem-solving, communication, and self-regulated learning. The study employed a systematic literature review supported by qualitative thematic synthesis. Searches were conducted in Scopus-indexed sources, ERIC, ScienceDirect, SpringerLink, MDPI, DOAJ, and Google Scholar using combinations of the terms “generative AI”, “ChatGPT”, “large language models”, “adaptive learning”, “personalized learning”, “student learning outcomes”, “digital literacy”, and “AI ethics”. Literature published between 2019 and 2025 was prioritized. Inclusion criteria covered peer-reviewed journal articles, systematic reviews, conference papers, and authoritative institutional reports directly related to GenAI, AI in education, adaptive or personalized learning, and student competency development. The final synthesis corpus consisted of 24 relevant sources after screening by title, abstract, full-text relevance, publication credibility, and methodological clarity. The findings show that GenAI supports adaptive learning through content personalization, immediate feedback, differentiated tasks, formative assessment support, learning reflection, and independent study assistance. However, its educational value depends on teacher supervision, clear AI-use policy, source verification, data protection, equitable access, and assessment designs that measure learning processes rather than AI-generated products. This review concludes that GenAI should be positioned as a pedagogical support system, not a replacement for teachers or students’ cognitive effort.