Generative AI in Secondary Language Education: Pedagogy, Skills, Outcomes, and Teacher Mediation

Generative artificial intelligence (GenAI) is increasingly used in language education, yet evidence from secondary-school settings remains fragmented. This integrative review examines empirical research on GenAI in formal secondary language education published from January 2023 to September 2026, with attention to pedagogical applications, language skills and activities, learning outcomes, teacher mediation, implementation challenges, and gaps in the evidence. Guided by Whittemore and Knafl’s integrative-review methodology and a PICo framework, the review drew on Google Scholar, ERIC, publisher and journal websites, and citation chasing. Twenty-two empirical studies were examined against the review criteria and available original publication records. Narrative and thematic synthesis were used to compare study contexts, methods, reported outcomes, and limitations. The studies describe uses of GenAI for writing support, feedback, conversation, tutoring, learning companionship, teacher support, and AI-literacy activities. Writing is the most represented language focus. Reported outcomes are mixed, ranging from gains in selected skills and learner perceptions to null or qualified findings. Overall, the evidence indicates that the educational contribution of GenAI depends on task design, teacher mediation, verification, AI literacy, and the conditions in which the technology is used.