Artificial intelligence in EFL vocabulary learning: a systematic review of higher education practices and research trends (2015–2025)

Vocabulary development is central to academic literacy in higher education EFL contexts, and the rapid expansion of artificial intelligence (AI) has intensified research in this domain. Following PRISMA 2020 guidelines, the present systematic review synthesizes 25 empirical studies (2015–2025) to map publication trends, examine methodological characteristics and forms of AI integration, and identify reported dimensions of vocabulary learning and associated outcomes. Research activity increased sharply after 2023, with quantitative and mixed-methods designs dominating the evidence base and AI integration expanding across diverse instructional arrangements. Reported dimensions of vocabulary learning focused primarily on breadth, whereas depth, delayed retention, and transfer outcomes were examined less frequently. Across studies, stronger outcomes tended to be reported when AI use was embedded in coherent pedagogical designs with explicit task alignment, structured scaffolding, and teacher mediation rather than implemented as stand-alone self-access support. The synthesis provides an updated overview of AI-supported vocabulary learning in higher education EFL contexts, highlights implementation features associated with more robust outcomes, and identifies priorities for stronger theoretical grounding and clearer alignment among objectives, tasks, and measures.