How eTwinning scaffolds pedagogical transformation through generative AI: a European cross-national study

While European policy agendas increasingly promote GenAI in schooling, less is known about how international professional communities support teachers in developing GenAI-related pedagogical competence. This qualitative study explores how participation in an AI-literacy eTwinning project shaped primary teachers’ GenAI integration across Greece, Romania, Slovenia, and Türkiye. Semi-structured interviews with eight teachers were analysed through thematic content analysis. Findings suggest that eTwinning functioned as a cross-national community of practice, enabling teachers to build confidence and pedagogical routines through distributed expertise, peer modelling, and sustained collaboration. Teachers most consistently used GenAI for differentiated instruction, rapid adaptation of learning materials, and multimodal creative production. However, trajectories varied by context, constrained by uneven infrastructure, limited institutional support, and policy ambiguity, pointing to an aspirational – infrastructural disjuncture. Teachers demonstrated strong everyday ethical vigilance, yet reported limited algorithmic understanding. Implications are discussed for scalable, community-based teacher professional development that integrates pedagogical practice with deeper GenAI literacy.