Artificially Intelligent systems have greatly modified digital learning and given rise to Intelligent Tutoring Systems (ITSs) which have the ability to provide personalized feedback at a large scale. Today's ITSs combine methods of adaptive learning, natural language processing, large language models (LLMs), and multimodal content delivery to cater to the diverse needs of learners. This survey reviews the fundamentals of ITS, shift from data-driven and machine-learning adaptivity, the coming of age of conversational AI tutors, and contemporary advancements in multimodal instructional technologies. Major research trends, system taxonomies, assessment methodologies and unresolved challenges relevant for AI-enabled tutoring are highlighted. The survey aims to give an overview of AI in transforming tutoring systems collectively while outlining opportunities for fresh research.
Venugopal et al. (Sat,) studied this question.