Shiyu Wang, Ying Cui, Xiaoquan Pan
Introduction As learners of English as a Foreign Language (EFL) frequently experience foreign language anxiety (FLA), a significant negative affective factor that adversely impacts learning outcomes warrants scholarly attention. Recent trends show an increasing integration of artificial intelligence (AI) in language education to prompt investigation into AI’s effectiveness on EFL learners’ foreign language anxiety. Nevertheless, systematic reviews synthesizing empirical findings on the effects of AI’s applications on EFL learners’ FLA remain scarce. Methods This study employed the PRISMA framework to conduct a comprehensive review of peer-reviewed empirical research published between January 2016 and December 2025. Results Based on systematic analysis, the findings are categorized into four dimensions: general characteristics of the researches, AI affordances and effectiveness, possible mechanism, other contributing factors. In addition, the factors influencing AI’s impact on learners’ FLA can be primarily synthesized into two sides, internal factors (learner characteristics, perceptions) and external factors (subjective elements and objective elements). Discussion This review thereby elucidates possible potential pathways through which AI influences EFL learners’ foreign language anxiety. Finally, the implications highlight the need for careful consideration of factors affecting AI implementation efficacy in language education. However, the proposed possible mechanism and influencing factors remain provisional and still require empirical testing in future studies.