Generative AI in Education: Transforming Personalized Learning

Abstract

Generative AI (GenAI) is increasingly discussed as a way to bring personalized learning to more students without multiplying the workload of already-stretched teachers. This paper looks at what that promise actually requires: not just a capable language model, but a human-centered tutoring framework that ties together learner data, curriculum-grounded content, adaptive generation, ongoing assessment, and teacher oversight. Drawing on recent empirical research, we find genuine reason for optimism about learning outcomes, but also recurring risks – hallucination, bias, privacy exposure, over-reliance, and academic-integrity concerns – that any deployment has to take seriously. We use these findings to work through applications, evaluation approaches, current limitations, and where the research still needs to go.