Abstract
Speaking is often the most demanding part of learning English in settings where students have few opportunities to use the language beyond the classroom. Artificial intelligence (AI) offers new ways to address this difficulty through voice-enabled conversations, role plays and immediate feedback. This article examines how AI-supported activities may contribute to speaking development in English as a foreign language (EFL) education. A narrative review of relevant theoretical work, systematic reviews, empirical studies and international guidance is used to consider oral practice, fluency, pronunciation, speaking anxiety, learner autonomy and the reliability of feedback. The literature suggests that conversational AI can provide valuable opportunities for rehearsal and may improve learners’ confidence in some contexts. These benefits are not uniform: they depend on the task, the learner’s proficiency, access to suitable technology and the quality of automated responses. Concerns about inaccurate corrections, accent recognition, privacy and dependence on prepared answers also deserve attention. Drawing on these findings, a five-stage teaching cycle is proposed in which students practise with AI, consider selected feedback and then apply what they have learned in conversation with other people. The article argues that the greatest value of AI lies in extending purposeful practice while keeping teachers and authentic communication central to speaking instruction.
This work is licensed under a Creative Commons Attribution 4.0 International License.