Master’s Dissertation Writing In The Chatgpt Era: Redesigning Policy And Assessment In Higher Education

Abstract

The integration of generative AI into academic writing has introduced new complexities in supervision, assessment, and the ethical dimensions of research production. This study investigates how EFL Master’s students use generative AI (Gen AI) tools during the dissertation writing process and examines the cognitive and academic consequences of such use. It also explores potential solutions to the challenges posed by Gen AI, drawing on insights from both students and teachers. Data were collected through student and teacher questionnaires and supervision-based reports. The participants included 52 Master’s students who defended their dissertations during the June–July 2025 session, and 15 teachers who supervised and/or examined these dissertations. Findings reveal that Gen AI tools were employed at nearly every stage of the dissertation process, including tasks traditionally expected to be performed by students themselves, such as summarizing literature and discussing findings. This extensive reliance led to cognitive offloading, which negatively impacted students’ intellectual development and critical thinking. Teachers also expressed academic concerns, often disapproving of work they perceived as overly AI-dependent. Students, in turn, acknowledged limitations in using Gen AI, such as AI hallucinations and the loss of personal voice and writing style. The study further highlights the absence of clear guidelines on the ethical use of Gen AI in academic writing and the lack of effective assessment methods for distinguishing between human- and machine-generated content. To address these challenges, the findings call for the inclusion of core principles such as transparency, responsible AI use, and a balance between human effort and machine assistance. In addition, AI-resistant assessment strategies are recommended, including enhanced questioning during oral defenses and greater reliance on supervisors' reports that track students’ progress over time. This study contributes to the evolving discourse on AI in higher education by emphasizing the need for responsible integration and strong, integrity-focused assessment practices