Exploring the integration of artificial intelligence in math education: Preservice Teachers’ experiences and reflections on problem‐posing activities with ChatGPT

Abstract This study addresses the need for research that incorporates educational perspectives and theories to understand the impact of Artificial Intelligence (AI) on learning and teaching. Utilizing AI‐integrated mathematical problem‐posing (AIM) activities and the AI‐powered scaffolding (AIS) strategy, the research investigated preservice teachers’ (PTs’) experiences with AI and their capacity for error recognition and correction in problems posed by ChatGPT. The findings reveal that while the PTs excelled at verifying error‐free problems, they struggled significantly with identifying and correcting errors, indicating a gap in their instructional preparedness. The study demonstrates that AIM activities are effective tools for assessing and developing PTs’ error recognition and correction skills. Additionally, AIM activities support the transfer of mathematical knowledge to pedagogical and instructional practices, contributing to PTs’ growth as adaptable educators. The research highlights the need to integrate AI‐based activities into PT training to build robust mathematical knowledge and teaching skills. Focusing on learning, pedagogy, and the human aspects of technology use, AIM activities and the AIS strategy enable PTs to engage critically with AI outputs and enhance their metacognitive skills. These insights emphasize the importance of incorporating AI‐integrated methods into teacher preparation programs to better equip future educators for an AI‐driven educational landscape.