Between Technical Assistance and Authorship: Researchers’ Ethical considerations of Using Generative AI for Scientific Writing

Abstract

Scientific writing is essential for sharing research findings, yet it poses significant challenges due to the need for precision, clarity, and adherence to strict standards. Generative AI (GenAI) tools, such as ChatGPT, have emerged as potential aid in this process, offering capabilities in linguistic editing, translation, and summarization. However, the integration of GenAI in academic writing raises ethical concerns regarding plagiarism, data fabrication, authorship and the preservation of the researcher’s original voice. This study empirically investigates researchers' ethical considerations regarding the use of generative AI in scientific writing. Through a survey of 102 researchers and research students in Israeli academia, the study explores the perceived ethical boundaries and conditions for appropriate GenAI use. Findings indicate that while GenAI can be beneficial for technical tasks, its role in content creation remains contentious. Transparency, critical evaluation of GenAI outputs, and maintaining originality are highlighted as essential conditions for ethical AI integration.