Artificial Intelligence in Education (AIEd) differs from other application areas of AI by having a vulnerable user group supposed to benefit from the research in the area and by frequently asking more of machine learning (ML) models than only to produce predictions. The first difference is apparent in the comparatively strong research tradition of AIEd concerning fairness and explainability. The second difference is pronounced in the motivation for AIEd research and tools; e.g., the aim is not to predict who is likely to drop out but to intervene to avoid dropout. In this paper, we argue that these differences lead to causal thinking being particularly valuable for AIEd compared to other application areas of artificial intelligence and argue in favor of establishing a tradition surrounding causal thinking. We state that a focus on causality can further increase the area’s existing strengths while simultaneously leading to new insights. We discuss that by incorporating causal thinking, AIEd becomes more accurate, fairer, more explainable, achieves better interventions, makes better recommendations, and can even make a more convincing case for its importance. Moreover, establishing a tradition around causal thinking naturally fosters interdisciplinary research and the inclusion of stakeholders such as teachers and students.