Researchers at Jefferson should make informed choices about using generative AI for their writing projects. Using AI tools ethically in scholarly publication is critical; using the tools and insights in this guide and adhering to the target publication’s policy can help. Many journals have already embedded AI into their authorship policies, which can be accessed through the instructions for authors on their websites. Adhering to their guidelines not only improves the chance of publication, but it also ensures that the scholarly work remains ethical and avoids issues of plagiarism.
Can Generative AI be an author?
Authorship for scholarly publication has several rules, including taking accountability for the work. Because AI tools cannot be accountable for a scholarly article, they cannot be considered authors of a publication. Still, if authors are using these tools to assist with generating scholarly articles, they must acknowledge that use within the paper. The APA Style Guide for ChatGPT (link below) offers specific guidance on how to do so ethically, including by describing the use of the tool in the Methods section of the manuscript, referencing the tool in the manuscript’s References section, or providing the full text of ChatGPT responses in an Appendix.
If you are using AI, check all your references and citations to ensure they maintain scholarly integrity. Generative AI has been known to create fake sources.
Further Reading:
What about AI and Intellectual Property?
Authors should also be aware that the use of AI in their scholarly work may create issues with Intellectual Property (IP).
Can authors use Generative AI?
We always advise authors to check the guidelines for the publication, conference, or grant. Keep in mind that your scholarly work may contain confidential material that your team may not want to become part of a large language model. This University of Nevada-Reno library guide provides a useful resource for authors, including evaluating sources and copyright.
Can peer reviewers use AI?
The short answer is—no. We caution against using AI for two reasons. First, entering a manuscript into AI can breech the confidentiality of the authors’ work. Second, the NIH prohibits the use of AI for peer review (June 2023).