NEGOTIATING IDENTITY AND INVESTMENT THROUGH AI-MEDIATED ENGLISH COMMUNICATION: EVIDENCE FROM INDONESIAN EFL LEARNERS

Abstract

Despite the rapid proliferation of large language model (LLM)-based conversational AI tools in educational settings, their implications for learner identity remain insufficiently examined, particularly in postcolonial multilingual contexts such as Indonesia.  This study aims to address this gap by examining how Indonesian EFL university students negotiate identity and investment in AI-mediated English communication through Norton's identity and investment framework. Data were collected through semi-structured interviews and participant-shared AI interaction logs from 12 undergraduate students at a private university in Central Java, Indonesia, and analysed using reflexive thematic analysis. It explores how learners position themselves as English speakers, how their investments in English shape engagement with AI, and what ideological tensions emerge from these interactions. The analysis identified four interrelated themes: imagined identity and aspirational investment, affective positioning and the non-judgmental interlocutor, contested legitimacy and native-speaker ideology, and agentive appropriation and its structural limits. The findings suggest that AI-mediated interaction functions as a site of identity negotiation, enabling learners to rehearse imagined futures, manage affective risks, and negotiate linguistic legitimacy. These findings contribute to emerging scholarship on AI-mediated language learning by illustrating how identity, investment, and agency are negotiated in AI-mediated English communication and by identifying implications for EFL pedagogy in Indonesian and comparable Global South contexts.