AI-Driven Education Systems: A Review of Applications, Explainable AI, and Generative Models

Over the past few years, Artificial Intelligence (AI) has grown significantly in modern education, and that has opened up further opportunities for learning analytics across various educational domains, intelligent tutoring and assessment systems as well as datafied personalization. In the last few years, some surveys and systematic reviews are starting to demonstrate how educational systems based on artificial intelligence are evolving from experimental prototypes to practical applications with a clearer impact in higher education and online learning [1], [2]. With such advancements, it is even more important to take a closer look at Explainable Artificial Intelligence (XAI) that targets explanation of AI-based decisions for educators and students [3, 4] beyond the trust and transparency but also focusing on fairness. More immediately, the rapid adoption of generative AI and large language models has accelerated integration of AI in education by making it easier to generate content, interactively tutor students, or provide personalized feedback -- while simultaneously raising questions about responsible use and academic integrity [5], [6]. Focusing on survey research, real-world applications, XAI based educational systems and generative AI for learning solutions, this paper provides a focused review of the AI in education literature dated from 2020 to 2025. To this end, in this review we discuss current challenges and open research directions to facilitate transparent, ethical and human-centric AI use in education by channeling key focus points on research trends, application areas as well as methodological paradigms.