Editorial Note: GenAI for Academic Writing – Friend or Foe?

In an article published in Nature titled “ChatGPT listed as author in research papers” Stone-Walker (2023) shocks the academic community with the fact that GenAI tools such as ChatGPT and Gemini have gained a substantive role in the production of knowledge and academic writing. He reports that one research company has published 80 articles produced by GenAI in academic journals. In the wake of Stone-Walker’s article, many publishers and journal editors set guidelines in relations to the role of GenAI in academic writings. All of them disagree to allow GenAI as an author. Further, the Stanford University Artificial Intelligence Index (2022) reports that there is a fivefold increase in research and publications on fairness and transparency relating to GenAI since 2014 indicating that the ethical issue is even more pressing now. Altogether, such development demonstrates that the academic community is feeling uneasy, disturbed, and anxious on the use of GenAI in the academic endeavour. Although everyone agrees on the practical assistance GenAI provides in academic writing, GenAI also brings epistemic challenges and accompanying integrity risks (Chesterman & Chieh, 2026). As a journal concerned with human behavior and socio-cultural processes in Asia, Makara Human Behavior Studies in Asia has a particular stake in addressing this issue as we take active roles in preserving academic authority related to journal publications. It is the aim of this editorial note to discuss principles in relation to how GenAI may be used in manuscripts submitted to this journal without sacrificing academic integrity. This editorial note does not yet introduce formal rules or technical instructions. Instead, it articulates the principles that will guide subsequent editorial policies. For this editorial note, GenAI refers to the term generative AI, which are computational techniques that are capable of generating seemingly new and meaningful content such as text, images, or audio from training data. (Feuerriegel et al., 2024). This can be used to perform tasks such as pattern recognition, prediction, generation, and optimization across research workflows.