Ethical Implications of Using Generative AI in English Language Teaching: Teacher and Student Perspectives

Abstract

This study investigates the ethical implications of integrating generative artificial intelligence into English language teaching, foregrounding the lived perspectives of both educators and learners. Drawing on qualitative data gathered from semi-structured interviews, classroom observations, and reflective journals across six diverse educational settings---including public secondary schools, university language centers, and private language academies---the research identifies three interlocking ethical domains: agency and authorship, equity and access, and pedagogical integrity. Teachers consistently expressed concern over diminished student ownership of language production when AI tools were used uncritically for writing or speaking practice, while students articulated ambivalence---valuing efficiency and personalized feedback yet reporting diminished confidence in autonomous expression. A recurring tension emerged between institutional pressure to adopt AI-enhanced platforms and teachers' professional judgment about appropriate scaffolding and assessment transparency. The findings reveal that ethical risks are not inherent to the technology itself but are activated through specific implementation patterns: opaque prompt engineering, unacknowledged AI mediation in graded tasks, and absence of co-developed classroom norms around tool use. Crucially, ethical awareness was found to be highly context-dependent---shaped by curriculum mandates, infrastructure constraints, and prevailing conceptions of language learning as either skill acquisition or identity formation. The study concludes that ethical integration requires neither blanket prohibition nor unconditional adoption, but rather deliberate, collectively negotiated practices grounded in linguistic epistemology and relational pedagogy.