This study examines EFL students’ perceptions of ChatGPT use in thesis writing, particularly in negotiating the boundary between academic assistance and cheating. Using a qualitative narrative inquiry design, data were collected through in-depth interviews to explore students’ experiences. The findings show that ChatGPT is perceived as a valuable academic tool that enhances linguistic accuracy, supports idea generation, and improves writing efficiency, especially for overcoming language barriers. However, its use also creates ethical ambiguity, as students often struggle to distinguish acceptable assistance from academic misconduct. Although students attempt to regulate their use of ChatGPT, their self-regulation is frequently inconsistent and influenced by contextual factors such as academic pressure. These findings highlight the need for clear institutional guidelines and the integration of AI literacy in academic writing instruction to promote responsible and ethical use of generative AI in higher education.
Copyrights © 2026