This study explores the integration of ChatGPT as a co-designer in mathematics education, focusing on its use by mathematics specialists and teachers to support culturally responsive pedagogy and differentiated instruction. As classrooms become increasingly diverse, educators face the dual challenge of designing lessons that are both inclusive and rigorous. Drawing on data from a graduate-level mathematics education course, this research examines how in-service teachers and mathematics specialists engaged with ChatGPT to co-create instructional tasks that reflect students' cultural backgrounds and support varied learning needs. Two structured assignments—one focused on culturally relevant task design and the other on differentiated instruction—provided the context for analyzing participants' reflections, AI-generated outputs, and collaborative lesson planning. Findings highlight the affordances of ChatGPT as a time-efficient idea generator and a tool for expanding teachers' instructional repertoire. However, participants also identified limitations, including mathematical inaccuracies, superficial cultural references, and misaligned task sequencing. Participants demonstrated strong curriculum and instructional design skills by critically evaluating and adapting AI-generated content, drawing on their specialized knowledge to ensure coherence, accessibility, and rigor. The study provides evidence of the indispensable role of teacher expertise in guiding the effective use of AI in education. An important implication for researchers and educators is the need to cultivate critical digital literacy to evaluate and adapt AI-generated instructional materials, ensuring that mathematics instruction is mathematically accurate, culturally authentic, and developmentally appropriate, while also leveraging collaborative professional learning to define AI's role in shaping equitable and coherent mathematical learning experiences.
Ⅰ. INTRODUCTION
Ⅱ. THEORETICAL BACKGROUND
Ⅲ. METHODS
Ⅳ. RESULTS
Ⅴ. DISCUSSION AND IMPLICATIONS
CONFLICTS OF INTEREST
REFERENCES
(0)
(0)