The careful observer of teaching methods at education and higher education levels in the Arab world (levels commonly known as key stages 1 to 5 in the UK), will quickly realise that the methods have not changed much over the decades, i.e. since the start of the twentieth century. Teacher-centred education is the norm with the classical use of the blackboard (or whiteboard) and, during the last quarter of century or so, the data show, the presence of a teacher being definitely important, in fact crucial, in providing proper guidance to the students. Nevertheless, in more crowded classrooms, especially in higher education amphitheatres, personalized attention to the student’s learning becomes very difficult, not to say impossible. A lot has been researched and written about learner-centred education through which learning is meant to improve as a result of the learner’s active request and participation. This learning includes the material content as well as an automated evaluation of the learner’s acquired knowledge and skills. This takes one straight into the realm of Adaptive Learning Systems (ALSs). It has been shown that this new paradigm can help many types of students improve on their own. If learner-centred education is not to replace the presence of a teacher, it can definitely be a very enriching complement that can make up for some of the shortcomings of teacher-centred learning. A few years ago, we started to research the development of a learner-centred digital environment to enable learners to improve their skills in the Arabic language in terms of vocabulary learning, reading comprehension, essay writing, and question-answering. The adopted approaches and techniques involve Artificial Intelligence (AI), Natural Language Processing (NLP), and machine learning. The aim of these prototypes is to eventually build comprehensive, customised tools for learner-centred education from Key stage 1 to Key stage 5 (i.e. till higher education) with the domains covering, for a start, the Arabic language and literature, Islamic Sciences, History and Geography. Recent developments in AI have produced a fairly revolutionary technology, the Large Language Model (LLM), best exemplified with the launching on the 30th of November 2022 of ChatGPT by OpenAI. This was the first time an LLM was introduced to the public at large, not just to be used by AI practitioners. Since then, numerous LLMs have been trained and made available to the public, either freely—sometimes integrated into search engines—or via a paid fee. In this chapter we present some of the work that has already been done on the use of LLMs in developing ALSs for reading comprehension of Arabic texts, development of a learner’s general culture (in Arabic literature, Islamic Sciences, history, geography, and science), learning of Arabic grammar, learning of English by Arab natives, etc. We also draw a roadmap of the future work that we envisage doing in order to productively introduce AI into the education and higher education contexts. This work is still prototypical in the sense that the ultimate aim is to see the same techniques generalized and applied in a national context and, indeed, in a Pan-Arab context.