World language teachers' deliberative engagement with generative AI for designing meaning‐based L2 reading instruction

This interpretive qualitative case study examines how in‐service world language teachers deliberatively used, selectively adopted, or avoided generative artificial intelligence (GenAI) tools (e.g., ChatGPT, MagicSchool) while designing meaning‐based L2 reading instruction with authentic texts. Grounded in a sociocultural view of teacher learning and informed by an interactive approach to reading, teachers' engagement with GenAI is interpreted as a socioculturally mediated process situated within a shared pedagogical framework. Eight in‐service teachers enrolled in a graduate‐level literacy methods course designed L2 reading lessons using PreGAP, a meaning‐based framework for designing L2 literacy instruction. Analysis of whole‐group discussions revealed substantial variability in GenAI engagement, ranging from deliberate adoption to principled avoidance. Teachers' decisions reflected framework‐mediated pedagogical reasoning grounded in pedagogical content knowledge (PCK) and technological pedagogical content knowledge (TPACK). Across participants, GenAI was most often positioned as a bounded thought partner rather than a lesson generator within the constraints of meaning‐based instructional design. Findings suggest that teacher preparation programs should prioritize rigorous, sustained engagement with interactive, meaning‐based frameworks for L2 reading instruction with authentic texts before introducing GenAI as a tool to assist with lesson design.