Vocabulary knowledge is a fundamental component of language proficiency and plays a crucial role in developing communicative competence in English as a Foreign Language (EFL) contexts. The rapid advancement of artificial intelligence (AI) technologies has significantly transformed the landscape of vocabulary learning and teaching. This systematic review critically examines empirical research published in Social Sciences Citation Index (SSCI) journals between 2020 and 2025, with a particular focus on the application of AI in EFL vocabulary instruction. To ensure methodological transparency and replicability, the review followed the PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) framework and employed a structured search process. A total of 32 empirical studies met the inclusion criteria and were analyzed in terms of participant profiles, vocabulary focus, types of AI technology employed, methodological design, and pedagogical outcomes. The findings reveal a diverse range of AI applications across the studies, including generative chatbots, automated feedback systems, object recognition tools, and game-based learning environments. Despite this diversity, several limitations were identified. First, the majority of studies employed experimental or mixed-methods designs, underscoring the need for more qualitative research to generate in-depth, context-sensitive insights into learners’ experiences, attitudes, and cognitive engagement with AI-assisted vocabulary learning. Second, the participant samples were predominantly composed of university students, limiting the generalizability of findings to younger or more diverse learner populations. Notably, no study addressed vocabulary learning within English for Specific Purposes (ESP) contexts, highlighting a significant gap in domain-specific instruction. Third, although many studies reported improvements in vocabulary use, retention, or lexical diversity, relatively few were explicitly designed to examine vocabulary instruction as a primary objective. In most cases, vocabulary outcomes were treated as secondary to broader goals such as writing, speaking, or general language development. Despite these limitations, the review consistently demonstrates that AI-enhanced environments can promote increased lexical diversity, provide personalized vocabulary feedback, and foster greater learner engagement. It underscores the importance of addressing underrepresented learner populations, adopting qualitative methodologies, and integrating vocabulary instruction into evolving pedagogical practices. The findings of this review offer critical implications for language educators, curriculum designers, educational technology developers, and researchers aiming to implement or investigate AI-supported vocabulary instruction across diverse EFL learning contexts.

Keywords: Artificial intelligence, vocabulary instruction, English as a Foreign Language, systematic review.