The impact of AI chatbots on English language learners’ speaking proficiency: A systematic review

Artificial intelligence (AI) chatbots have become a significant phenomenon not only in the information technology field but also in education, where they are transforming how students learn English. Numerous studies have explored the use of AI chatbots in English language teaching (ELT), especially their potential to support speaking practice. However, the existing research is fragmented and lacks a systematic synthesis, making the findings inconclusive about how chatbots contribute to speaking proficiency and how learners perceive their use. This limitation creates uncertainty for educators, policymakers, and researchers in assessing the educational value of chatbots in ELT. To address this gap, this study aims to critically analyze how AI chatbots enhance students' English-speaking skills and the implications for developing speaking proficiency, as well as investigate students' perceptions of their effectiveness. This systematic literature review examines findings from various studies related to AI chatbot applications in English-speaking educational settings, published between 2018 and 2025. Using the PRISMA guidelines, an initial search across Google Scholar and Science Direct yielded 111 papers; the findings reveal both benefits, such as increased learner engagement and opportunities to practice, and challenges, including the lack of accuracy. By synthesizing prior research, this study offers a clearer understanding of the role of chatbots in increasing speaking proficiency and contributes to ongoing debates about AI in ELT.