A Systematic Review on Generative AI in L2 Speaking Development From a Self-Determination Theory Perspective

Abstract

Second or foreign language (L2) speaking is a key aspect of L2 competence. While GenAI shows potential to address challenges in L2 speaking development, its role remains insufficiently understood. Self-Determination Theory (SDT) could explain student language learning engagement. Therefore, this systematic review examines how GenAI affordances can support learners’ SDT needs satisfaction and, in turn, influence their behavioral, cognitive, and emotional engagement during GenAI-assisted L2 speaking development. We selected 30 relevant empirical studies through systematic searches of ProQuest, Web of Science Core Collection, and Scopus. Our findings reveal that GenAI notably enhances autonomy support and substantially satisfies the need for competence, although it does not fully address relatedness satisfaction. Furthermore, needs satisfaction influences learners' behavioral, cognitive, and emotional engagement (or disengagement) in GenAI-supported L2 speaking development at varying levels. These findings extend the SDT literature in GenAI-assisted learning and offer empirical and practical insights for the effective integration of emerging technologies into L2 education.