This review explored how intelligent systems are when used in educational settings to assist in tutoring. Their effectiveness, the different methods used, and their limitations were analyzed. A screening of 30 articles was carried out by the researchers, identifying 16 of them between 2018 and 2024. The results show that various technologies, including intelligent tutoring systems (ITSs), Smart MOOC integrated into Intelligent Tutoring (SMIT), fuzzy Bayesian intelligent tutoring systems (FB-ITSs), as well as AI tools like ChatGPT and Meta-Tutor, are being utilized. These systems improved personalized learning by providing adaptive and interactive educational environments that can change traditional teaching methods. It was also found that they have significantly improved how engaged students are, how well they do in their studies, and how customized their learning experiences are. However, some crucial drawbacks were identified, such as how AI can’t quite match the subtle, human parts of teaching, like providing emotional support and helping students grow. To address these challenges, it is suggested that future advancements in education. Artificial Intelligence systems should be developed to improve or support human abilities, not necessarily replace them. Considering such matters as confidentiality, fairness, or discrimination in machine learning models is equally significant and fundamental. These systems require an infrastructure that precedes issues related to technology development and values while ensuring equitable participation at different stages of learning.