Intelligent tutoring systems were one of the early uses of AI, in particular trying to adapt content to the level of attainment and modes of learning of a pupil. In the early days, this was most useful for formal topics such as mathematics, but LLMs have transformed this allowing much more personalised instruction in most areas of the curriculum, although with the need for clear guards, to prevent inaccurate or innappropriate responses. Closely related are learning analytics, which are usually more about asssessing overall progress, but the two overlap, for example if the learning analytics suggest a student has problems in a certain area this can be used to suggest additional remedial or revision material.