Bridging or breaking? A systematic review of how generative AI shapes equity in foreign language education

Abstract

This systematic review examines how generative AI (GenAI) technologies, such as ChatGPT, shape equity in English as a Foreign Language (EFL) education. Drawing on 22 peer-reviewed empirical studies published between January 2023 and June 2025, selected from an initial pool of 430 records across five major databases (Scopus, Web of Science, ERIC, ProQuest, and Google Scholar), the review synthesizes emerging evidence on GenAI's dual impact. Findings reveal that GenAI can bridge equity gaps by expanding access, providing personalized scaffolding, supporting multilingual and inclusive learning, and promoting teacher innovation, yet it can also widen disparities through infrastructural inequities, digital literacy divides, algorithmic bias, and English-dominant linguistic norms. To address these tensions, the review introduces the Equity-Positioned Implementation of AI (EPI-AI) Framework, which conceptualizes equitable GenAI integration as systemic, relational, and contextually contingent across five interrelated domains: access and infrastructure, AI literacy, pedagogy, governance, and sociocultural responsiveness. By consolidating empirical insights through an educational justice lens, this review repositions GenAI research in EFL from tool-focused innovation toward an evidence-based and equity-oriented understanding of how technology mediates inclusion, participation, and opportunity.