The Impact of AI-Supported Speaking Practice on EFL Learners’ Confidence in Vietnam

AI-supported speaking practice represents a transformative dimension of digital pedagogy by offering learners accessible, adaptive, and low-stakes opportunities for oral language development. This study explored its relationship with EFL learners’ confidence in Vietnam through a quantitative correlational design. A total of 120 Grade 5 students engaged in a four-week intervention combining AI-driven chatbots and pronunciation feedback systems with regular English instruction. Data collected from Likert-scale questionnaires and usage logs revealed a statistically significant moderate positive correlation (r = .46, p < .01), indicating that frequent AI practice was associated with greater self-reported confidence, perceived fluency, and willingness to communicate. Qualitative interviews corroborated these findings, with learners reporting that real-time, non-judgmental feedback and the flexibility to practice “anytime, anywhere” enhanced motivation and reduced speaking anxiety. These results resonate with recent evidence that digital and AI-enhanced platforms can foster affective gains when paired with structured pedagogical design. Overall, the study contributes empirical support to applied linguistics by demonstrating that AI tools, when integrated thoughtfully with teacher guidance, can strengthen learner confidence and inform curriculum innovation and policy development in low-exposure EFL contexts.