Exploring the Ethical Issues of Using ChatGPT in Language Learning among Iranian EFL University Students

This study investigates the ethical contours of ChatGPT-supported language learning among Iranian university students of English as a foreign language. Using semi-structured Persian-language interviews with ۱۸ participants and a reflexive thematic analysis, the research maps how learners negotiate privacy, fairness, authorship, and learning efficacy when engaging with generative AI. Students reported privacy-first practices (removing identifiers, avoiding uploads, preferring vetted tools), persistent anxiety about detector bias toward non-native writing, and a normative boundary between scaffolding and substitution, as evidenced in prompt logs and self-justifications. They valued rapid feedback yet warned that unexamined automation can erode critical thinking, motivation, and confidence, especially when outputs default to non-local registers that require deliberate “localization.” Participants called for process-visible assessment, plain-language data notices, vetted platforms, and due-process pathways for contesting automated accusations. They also highlighted infrastructural inequities—VPN-dependent connectivity and device constraints—that condition who can benefit from AI. Taken together, the findings portray students as strategic, ethically attentive users who seek conditions under which ChatGPT augments rather than displaces language work. The study contributes a context-sensitive account of ethics-by-design for EFL, clarifying how institutional governance and classroom routines can sustain augmented autonomy, fairness, and data stewardship in Iranian higher education. Student perspectives foreground practice over hype narratives.

کلیدواژه ها:

Academic Integrity ، ChatGPT ، Ethical Issues ، Iranian EFL University Students ، Language Learning