29.04.2026 – Prof. Dr. Mustafa Cevikbas (Vrije Universiteit Brussel)
Artificial Intelligence in Mathematics Education: Promise, Pitfalls, and Empirical Insights from A Mathematical Modelling Perspective
Artificial intelligence (AI) is rapidly reshaping mathematics education research and practice, raising new questions about teaching, learning, assessment, and teacher preparation. This talk offers a broad overview of current developments in the field and highlights emerging directions for future research. It begins by drawing on our recent narrative review to map the conceptual landscape of AI in mathematics education, with attention to major AI technologies, their educational uses, and the opportunities and risks they create for learners and teachers. In particular, the talk discusses the potential of AI to support mathematical reasoning, problem-solving, and adaptive learning, while also addressing concerns related to reliability, overreliance, equity, ethics, and teacher education.
The second part of the talk presents the main contributions of a recent special issue of ZDM on AI research in mathematics education, with a focus on how current studies are extending and diversifying the field. The talk concludes with findings from a recent empirical study on preservice mathematics teachers’ use of large language models in solving mathematical modelling problems. This study examines how participants used LLMs across modelling phases, which affordances and challenges they perceived, and how they judged the trustworthiness of AI-generated outputs. Taken together, the talk argues for a critical, conceptually grounded, and pedagogically responsible approach to AI in mathematics education.