AI-MEDIATED TRANSLANGUAGING IN MULTILINGUAL EFL CLASSROOMS: EXAMINING LEARNER ENGAGEMENT, AUTONOMY, AND LANGUAGE DEVELOPMENT IN THE PAKISTANI CONTEXT

Abstract

The rapid integration of generative artificial intelligence (GenAI) in education has transformed language learning practices, especially in multilingual contexts. This study investigates how GenAI mediates translanguaging in English as a Foreign Language (EFL) classrooms in Pakistan, exploring its impact on learner engagement, autonomy, language development, and inclusion. A convergent mixed-methods design was employed, involving 52 participants, including students and instructors, who engaged in AI-supported tasks such as chatbot-assisted discussions, AI-based writing, and guided reading exercises. Quantitative data from Likert-scale questionnaires were analysed using descriptive statistics, while qualitative responses were examined through thematic analysis. Findings indicate that AI tools facilitate the strategic use of learners’ full linguistic repertoires, enhancing comprehension, classroom participation, and collaborative interaction. Participants reported increased autonomy, reflective learning, and confidence in language production, alongside positive perceptions of equity and identity recognition in multilingual learning spaces. Despite minor challenges related to language switching and AI feedback accuracy, overall attitudes toward AI-mediated translanguaging were strongly positive. The study underscores the pedagogical potential of integrating AI and translanguaging to support inclusive, learner-centred EFL practices in Pakistan and similar multilingual contexts.