Exploring the Impact of AI Tools on English Language Teaching and Learner Autonomy

This paper examines the impact of Artificial Intelligence (AI) tools on English Language Teaching (ELT) and learner autonomy at the postgraduate level in higher education. Previous studies have highlighted the potential of AWE, ASR, and chatbot applications to improve specific language skills; however, there is a paucity of empirical research regarding their broader impact on students’ self-regulation and independent learning. To address this gap, a quantitative cross-sectional survey was conducted with a sample of 400 participants (300 students and 100 instructors) from higher education institutions utilizing AI-supported platforms. We used a structured questionnaire with a 5-point Likert scale to collect data on AI adoption, teaching and learning benefits, and learner autonomy. Reliability analysis was used to figure out how consistent the data was internally, and it showed that it was (r >. 80). Results demonstrated strong perceptions of the overall adoption of AI (M = 4.08, SD = 0.59), teaching positive outcomes (M = 4.12, SD = 0.55), and learner autonomy effects of AI as well (M = 4.02, SD = 0.61). Results of the ANOVA test showed that differences were significant according to proficiency levels, such that more advanced students (B2+) claimed to be more autonomous than intermediate (B1) and elementary (A2) students (F = 15.42; p <. 001). In addition, multiple regression analysis showed that AI adoption strongly predicted learner autonomy (β =. 33, p <. 001) and engagement (β =. 29, p <. 001), teaching (β =. 18, p <. 001) contributing moderately. These findings indicate that although performance plays a role, the increases in autonomy are largely a consequence of actively and meaningfully interacting with AI tools. The study concludes that AI has the potential to reshape ELT as a learner-centred practice and encourage self-regulation, autonomous practice, and communicative competence. However, it underscores the importance of access equality, teacher mediation, and ethical policy to achieve best results. Longitudinal and mixed-methods designs should be used in future work to investigate the development of the enhanced autonomy-enabled by AI.