Integrating AI Tools in Blended EFL Classrooms: Effects on Learner Autonomy and Performance

Abstract This study explores the impact of integrating artificial intelligence (AI) tools into a blended English as a Foreign Language (EFL) classroom on learner autonomy and language performance. Conducted at To Ky Secondary School in Ho Chi Minh City, the research employed a quasi-experimental mixed-methods design involving 80 students divided into an experimental group and a control group over a 10-week period. The experimental group engaged with AI-supported learning tools such as ChatGPT, Grammarly, and QuillBot, while the control group followed a traditional blended learning approach without AI integration. Data were collected through pre- and post-tests, questionnaires, interviews, and classroom observations. The findings revealed that students in the experimental group showed significantly greater improvement in both learner autonomy and language performance, particularly in terms of accuracy, fluency, and complexity. Additionally, qualitative results indicated increased motivation, enhanced self-regulated learning behaviors, and positive perceptions toward AI-assisted learning. However, challenges such as over-reliance on AI tools and concerns about the accuracy of generated content were also identified. Overall, the study highlights the effectiveness of AI-integrated blended learning in improving EFL outcomes and underscores the importance of guided implementation to ensure meaningful and sustainable learning.