Forming Policies for Ethical AI Use in Academic Writing

Easy access to generative AI has raised concerns about students’ engaging in academic dishonesty in their writing. AI in writing where a text written by an AI text generating software is used and submitted for dissemination or evaluation is presented as the students’ own work can be considered to be plagiarism. Students in higher education institutions can be penalized if AI text is detected in their examinations. However, as AI text generators can also be used to support and improve a student’s writing, there is a question as to what extent and what types of usage of AI text generators can be considered as plagiarism. Thus, it would be of great advantage to students if their respective higher education institutions can have clear guidelines and policies on AI use in students’ academic writing. This literature survey analyses the policies and instructions with regard to AI use in writing developed by the topmost universities in four English as a first language countries, namely Melbourne (Australia), Auckland (New Zealand), Cambridge (UK) and Princeton (USA). The literature analysis particularly focused on the two themes of (1) instructions to academic staff and (2) instructors to students. Significant instructions have been tabulated and discussed. They hold potential value as guidelines for consideration when formulating and improving policies for local state and non-state higher educational institutes. Keywords: AI text generators; ChatGPT; Plagiarism; Higher education policies