Exploring Undergraduate Students’ Experiences with Generative AI in English Language Learning

Generative artificial intelligence (GenAI) is increasingly embedded in higher education, yet its educational value depends not only on what these systems can do but also on how learners incorporate them into their learning practices. This study explores how undergraduate students experience and integrate GenAI into English language learning and how they negotiate its perceived benefits, challenges, and future role. A qualitative exploratory design was employed using an open-ended questionnaire completed by ۱۳۰ undergraduate students from Engineering and Foreign Languages and Literature programmes at a university in Iran. The data were analysed using Reflexive Thematic Analysis following Braun and Clarke’s approach. The analysis generated five interconnected themes: personalised language-learning support, AI-mediated scaffolding and interaction, the autonomy-dependence paradox, critical and metalinguistic awareness, and the changing ecology of English language learning. The findings indicate that students value GenAI for immediate feedback, personalised explanations, vocabulary development, writing support, conversational rehearsal, and low-pressure opportunities for practice. However, these affordances also create risks of reduced cognitive effort, dependence on immediate answers, reliance on AI-generated confirmation, and weakened confidence in learners’ own linguistic judgements. The findings suggest that GenAI can expand students’ procedural autonomy while simultaneously creating conditions for epistemic dependence. The study therefore advances an autonomy-dependence perspective in which the educational value of GenAI is determined by whether AI extends or displaces learners’ cognitive and linguistic activity. The findings further indicate that teachers remain central, although their role increasingly involves designing AI-mediated learning experiences, developing students’ critical AI literacy, and protecting learner agency. The study concludes that sustainable GenAI integration in English language education requires human-AI complementarity rather than technological substitution.

کلیدواژه ها:

English language learning ، generative artificial intelligence ، higher education ، AI literacy ، learner autonomy ، Reflexive Thematic Analysis