Student Perspectives on Ethical Academic Writing with ChatGPT: An Empirical Study in Higher Education

Abstract

The emergence of ChatGPT has significantly reshaped the landscape of higher education, sparking concerns about its potential misuse for academic plagiarism (Cotton et al., 2023). This study examines the use of ChatGPT in academic writing among students at the University of Mannheim in Germany and St. Gallen in Switzerland, using a proposed Human-AI collaboration framework with six levels of AI-enabled text generation (Boyd-Graber et al., 2023). The survey of 699 students reveals varied ChatGPT usage across all six levels, with Level 3 (Literature Search) being slightly more utilized. Students expressed mixed opinions on ethical issues, such as the declaration of ChatGPT-generated content in academic work and the extent to which ChatGPT is allowed at their university. The results of the study highlight students' concerns about negative effects on grades, a lack of clarity about university policies on ChatGPT, and fears that hard work will not be rewarded. Despite these issues, most students support open access to ChatGPT. The findings suggest the need for clear ethical guidelines in academia regarding AI use and highlight the potential stigmatization of AI, which could hinder technology acceptance and AI-related skills development.