Practical Considerations and Ethical Implications of Using Artificial Intelligence in Writing Scientific Manuscripts

Abstract

The growing accessibility and sophistication of artificial intelligence (AI) tools have transformed many areas of research, including scientific writing. AI tools, such as natural language processing models and machine learning-based writing assistants, are increasingly used to help draft, edit, and refine scientific manuscripts. However, the use of AI in the writing process introduces both legal and ethical challenges. Various guidelines and policies have emerged, particularly from academic publishers, aimed at ensuring transparency and maintaining the integrity of scientific work. This editorial aimed to provide guidance for researchers on the ethical and practical considerations regarding the use of AI in writing scientific manuscripts, focusing on institutional policies, authorship accountability, intellectual property concerns, plagiarism issues, and image integrity. INSTITUTIONAL AND JOURNAL POLICIES: TRANSPARENCY AND GUIDELINES As AI tools become more prevalent in scientific writing, academic institutions and journals are establishing clear guidelines regarding their use. The majority of well-known publishers provide guidance to authors, readers, reviewers, and editors concerning the role of AI-assisted technologies in the writing process, but industrywide standards have not yet been fully refined.1 The recently developed framework by American College of Gastroenterology (ACG) and Wolters Kluwer, the publisher of ACG journals emphasizes transparency and accountability, highlighting that although AI can assist in improving the readability and language of a manuscript, it should not replace the core tasks of authorship, such as generating scientific insights or drawing conclusions.2 ACG also requires authors to disclose the use of generative AI and AI-assisted technologies during the writing process. This disclosure fosters trust and ensures compliance with the terms of use of the AI tools. The goal is to provide clarity to readers, reviewers, and editors, helping them understand where and how AI was applied in manuscript preparation. Failure to disclose AI use can lead to ethical breaches, retractions, and potential damage to the professional reputations of both the researcher and the journal. The policy only covers the creation of new content and expressly forbids using AI on previously published material, preventing concerns around self-plagiarism or unauthorized content modification. AUTHORSHIP AND ACCOUNTABILITY: THE ROLE OF HUMAN OVERSIGHT The rise of AI in scientific writing raises fundamental questions about authorship and responsibility. The ACG's authors instructions emphasize that authorship cannot be attributed to AI, as AI cannot take responsibility for the accuracy or integrity of scientific work. Authorship implies a set of ethical and intellectual responsibilities that only human researchers can fulfill. Every listed author must be accountable for the content of the manuscript, and AI cannot be assigned such responsibilities. Moreover, although AI can assist in writing, it cannot generate the intellectual contributions to scientific research. Scientific manuscripts must be the product of human insight and critical thinking. ACG's policy highlights the need for human oversight in the application of AI technologies, emphasizing that all AI-generated content should be thoroughly reviewed and edited by the authors to ensure accuracy, completeness, and lack of bias. The authors are ultimately responsible for ensuring that their work adheres to the highest standards of scientific integrity. This policy aligns with broader academic norms, such as the International Committee of Medical Journal Editors criteria for authorship, which require authors to have made significant intellectual contributions to the research and to be accountable for the final work. Failure to ensure human oversight and control can lead to inaccurate or misleading scientific conclusions, jeopardizing the validity of the research. INTELLECTUAL PROPERTY AND OWNERSHIP OF AI-GENERATED CONTENT The use of AI in scientific writing introduces important questions about intellectual property (IP) ownership. Scientific manuscripts often contain novel ideas and discoveries, and researchers must be careful about the terms and conditions associated with the AI tools they use. Some AI platforms, such as OpenAI's Chat-GPT models, explicitly state that users retain ownership of the content generated through the tool. However, other platforms may have different terms, leading to potential conflicts over content ownership. ACG's policy ensures that authors remain responsible for the originality of their work, cautioning against the use of AI tools in ways that might lead to copyright violations or IP disputes. Because AI models are trained on vast data sets that may include copyrighted material, authors must be vigilant in ensuring that the AI-generated text does not unintentionally replicate existing works without proper attribution. This