AI Tools and Writing Identity in EFL Contexts: A Mixed-Methods Study of Voice, Ownership, and Cultural Authenticity

Artificial intelligence tools are increasingly integrated into second language writing (L2) classrooms, offering immediate feedback and scaffolding. While their potential to enhance accuracy and fluency is widely acknowledged, concerns remain about how AI mediation affects learners’ voice and identity. This study investigates whether first-year university students perceive a loss of authenticity in their writing when assisted by AI. A mixed-methods explanatory convergent design was employed with twenty first-year Algerian university students enrolled in English writing courses at the Affiliate of Teacher Education College-Tebessa. A structured questionnaire was used to focus on perceptions of voice, identity, and ownership in AI-assisted texts. Quantitative data were analyzed using descriptive statistics, while qualitative responses were coded thematically to capture nuanced reflections on authenticity and cultural expression. Findings indicate that students value AI for reducing anxiety and improving grammatical accuracy, yet many express ambivalence about authorship and personal voice. Patterns suggest selective acceptance of AI feedback, with students often re-editing outputs to reintroduce their preferred style. The study highlights the dual role of AI as both an enabler and a potential suppressor of identity in L2 writing. These findings underscore the need for pedagogical strategies that encourage critical engagement with AI suggestions.