Supporting Chemistry Teachers’ AI Integrated Practices: A Co Design–Based LUMA AI Online Forum Model and an AI PCK Framework

Artificial intelligence (AI) is increasingly embedded in chemistry teachers’ everyday practices. At the same time, AI generated chemistry content varies in accuracy, representational quality and pedagogical suitability, creating new demands for teachers’ professional judgement. This article presents two practice-oriented contributions designed to support chemistry teachers’ AI integrated practices. First, we introduce a co-design–based LUMA AI Online Forum model, grounded in the Evidence Based Teacher Education (EBTE) framework. The forum brought together chemistry teachers and teacher educators in sessions where AI generated chemistry representations, explanations and experimental suggestions were analysed, critiqued and redesigned through collaborative cycles. AI was approached as an object of pedagogical, disciplinary and ethical evaluation. Second, insights from the co-design process of the forum informed the development of an AI Pedagogical Content Knowledge (AI-PCK) framework. The framework builds on classical PCK by identifying how AI‑integrated chemistry teaching reshapes all five established domains – orientations toward chemistry teaching, curriculum knowledge, knowledge of students’ understanding, instructional strategies and representations, and assessment knowledge – while articulating AI literacy with ethical–epistemic competence as a necessary extension. Together, these contributions offer tools for strengthening chemistry teachers’ pedagogical judgement and professional agency in AI rich educational contexts.