AI-Powered Intelligent Tutoring Systems in Higher Education: A Review of Current Approaches, Challenges, and Future Directions

This paper presents a systematic review of ten recent studies examining the design and application of Intelligent Tutoring Systems (ITS) in higher education. The study explores the technological foundation of modern ITS, emphasizing personalization strategies, feedback mechanisms, and implementation challenges. Findings revealed a clear shift toward systems powered by Large Language Models (LLMs) and adaptive learning architectures to deliver a more personalized, scalable, and human-like learning experience and individual support. Despite the progress, ethical concerns, transparency, and scalability remain critical challenges.What sets this study apart is its timely synthesis of interdisciplinary work that bridges AI technologies with pedagogical application in higher education. Theoretically, it contributes by identifying emerging design patterns and categorizing evolving AI techniques within ITS. Practically, it offers insights for educators, institutions, and developers seeking to adopt or design ITS that are both context-aware and pedagogically sound. This review highlights key directions for future development and responsible implementation of AI in educational settings.