Adaptive Learning Systems Powered by Generative AI and Large Language Models: A Comprehensive Review of Personalized Education
Author1 Author 2
Ather Iqbal , Deven M. Kene
Assistant Professor Assistant Professor
Department of Computer Science Department of Computer Science
Vidya Bharati Mahavidyalaya Amravati. Vidya Bharati Mahavidyalaya Amravati.
atheramt13@gmail.com dkene75@gmail.com
ABSTRACT
Adaptive learning systems leveraging Generative Artificial Intelligence (GenAI) and Large Language Models (LLMs) are transforming personalized education by overcoming the limitations of traditional, one-size-fits-all learning approaches. Classical learning systems typically deliver static content and standardized assessments, offering limited flexibility to accommodate diverse learner abilities, learning paces, and engagement levels. In contrast, LLM-driven adaptive learning systems continuously analyze learner performance, behavioral patterns, and interaction data to dynamically personalize instructional content, adjust difficulty levels, and generate context-aware assessments and feedback in real time.
Recent advancements in LLMs enable natural language interaction, human-like tutoring, and adaptive scaffolding, allowing systems to identify knowledge gaps, address misconceptions, and support individualized learning pathways. The integration of predictive analytics further enhances adaptive learning by enabling early identification of at-risk learners and facilitating timely academic interventions to improve retention and learning outcomes.
This review synthesizes current research and practical implementations of generative-AI-powered adaptive learning systems across four key domains: AI-driven personalized learning, predictive analytics for student success, intelligent tutoring systems, and LLM-centric deployment architectures for web and mobile learning environments. The paper highlights system architectures, pedagogical benefits, and performance outcomes, while also discussing challenges related to data privacy, algorithmic bias, explainability, and scalability. Finally, it identifies future directions, positioning LLM-based adaptive learning systems as a foundational technology for next-generation, learner-centric education.
Keywords
Adaptive Learning Systems; Generative Artificial Intelligence; Large Language Models; Personalized Education; Intelligent Tutoring Systems; Predictive Learning Analytics; Educational Data Mining; Human-Centered AI
