ENGLISH-MEDIUM INSTRUCTION TEACHERS’ PERSPECTIVES ON AI- MEDIATED TRANSLANGUAGING: A CASE STUDY AT A VIETNAMESE UNIVERSITY

Abstract

Englishmedium instruction is developing quickly in Vietnamese universities as a result of national internationalisation efforts, but its implementation tends to encounter the multilingual nature of classrooms and the rise of new digital tools. This study, conducted between October 2025 and February 2026, explores how Englishmedium instruction lecturers at Can Tho University perceive and practise translanguaging, particularly how generative artificial intelligence begins to influence these practices in Englishfocused teaching environments. Drawing on semistructured interviews with five lecturers from different disciplines, the study indicates that Englishmedium instruction teachers frequently and intentionally employ both Vietnamese and English to help students grasp complex ideas, fully comprehend subject content, and participate more confidently, though still expecting English in formal academic tasks. From the participants’ perspectives, generative artificial intelligence appeared to support bilingual meaning-making, particularly when teachers and students moved between Vietnamese and English during explanation and task preparation. Teachers commonly prompt artificial intelligence in Vietnamese and then refine the output in English to enhance students’ comprehension and confidence, especially for students with lower English proficiency. Nevertheless, they also express concerns about shallow learning, uneven dependence on artificial intelligence, and the lack of obvious institutional guidance. Overall, the findings indicate that AI-mediated translanguaging may help teachers deal with some tensions between Englishmedium instruction policy and classroom realities in this context. The studyencourages future consideration of stronger institutional support, professional development, and assessment practices that appreciate multilingual and AIsupported learning.