This study systematically reviews existing literature and conducts a bibliometric analysis of Scopus data from 2016 to April 30, 2025, to explore current patterns and themes surrounding Artificial Intelligence (AI's) role in scholarly writing. The increasing volume of publications in this field underscores the growing importance of AI in academic writing. Our review highlights that “artificial intelligence”, “academic writing”, “ethics”, and “ChatGPT” are emerging as central topics in academic discourse, and that some scholarly articles have recognized ChatGPT as an author. The most contributing countries include Japan, Canada, Indonesia, the USA, India, and the Philippines, with substantial publications in reputable journals. The review reveals that while AI tools offer substantial advantages, such as enhanced writing support, critical brainstorming, streamlined research processes, and personalized assistance, they also pose challenges, including ethical concerns, potential inaccuracies, overreliance, bias, and monetary constraint. In addition, in scholarly writings, very often the AI-text generator is being used to generate manuscripts and, after that, humanized through other AI bypassing tools, which very often fail to be detected by AI-text detectors, questioning the effectiveness, reliability, and addressing the harmful nature of the AI tools in academic writing. Thus, future writing-dependent disciplines should use AI responsibly, verify AI-generated content, ensure data privacy and confidentiality, follow course-specific AI policies and acknowledge AI use when required. This research could be pivotal for various stakeholders, including researchers, educators, tech developers, policymakers, and students, providing insights into the adoption of AI in academic writing and addressing its potential benefits and drawbacks.
Ghosh et al. (Fri,) studied this question.