Students' Use of Generative AI for Academic Writing: Opportunities, Challenges, and Ethical Concerns

Abstract

This is a qualitative phenomenological study of the experiences of students in the rural secondary schools of Sunsari district on the use of generative artificial intelligence (GenAI) tools (ChatGPT, Google Gemini and Microsoft Copilot) in academic writing. Semi-structured interviews were carried out with eight purposively selected students from two rural secondary schools. Themes that emerged from the students' experiences with opportunities, challenges and ethical issues were identified using Braun and Clarke's (2006) reflexive thematic analysis. The results show that participants considered Gen AI useful for providing ideas, organizing the writing, expanding and correcting vocabulary and grammar, and simplifying complex explanations, which helped them gain confidence in academic writing. But there were several issues that arose, such as unreliable internet, access to personal devices, cost of mobile data, and different policies within schools. Concerns included academic honesty, unclear boundaries between acceptable AI support and plagiarism, and the fear that too much reliance on AI could make students dependent on it and reduce their ability to learn independently. The findings reveal a digital divide in terms of both access to technology and usage of GenAI tools in educational environments in rural areas. Students' personal standards around the use of AI were developing but there were some uncertainties about what is acceptable. The findings highlight the critical importance of delivering Gen AI digital literacy instruction in a way that is responsive to the context, creating and enforcing school-level policies for responsible use of GenAI, and providing pedagogical guidance for under-resourced rural schools to ensure successful and meaningful Gen AI use.