학부 유학생의 생성형 AI 활용 쓰기 윤리 인식 연구 -딜레마 기반 토의 활동을 중심으로-

Abstract

This study examines the ethical awareness of undergraduate international students regarding the use of generative AI tools for academic writing. As technologies such as ChatGPT and Bing AI become increasingly prevalent in higher education, students—particularly those writing in a second language—face complex dilemmas concerning authorship, plagiarism, and responsibility. To explore these issues, this study implemented a dilemma-based discussion model for international undergraduate students in a writing course at a Korean university. Using a mixed-methods approach, the study conducted a 12-item survey which was administered before and after the discussion sessions, and the results were analyzed using paired-samples t-tests and Cohen's d. The analysis revealed significant improvements in students' ethical awareness, especially regarding authorship and accountability for AI-generated content. Additionally, thematic analysis of the resarch participants' reflective journals identified four key themes: distancing themselves from AI dependency, the emergence of authorship identity, greater moral responsibility, and the internalization of academic integrity. The findings suggest that dilemma-based discussions foster ethical reasoning and critical digital literacy, offering a promising approach to AI-integrated writing instruction for diverse student populations.