Artificial intelligence (AI) is rapidly changing the teaching and research landscape of higher education. Policies and guidelines about the use of AI for research activities, specifically the writing of manuscripts and proposals, are being developed across agencies, publishers, and academic institutions. Investigators, project staff and students must become aware of the evolving policies on use of AI tools and question their reliability for use in the research environment.
Below are some Questions and Answers to consider but be aware that these may change abruptly as thing evolve. If in doubt, check the policies of the intended publisher or grant-making organization for up-to-date policies.
Can AI be listed as an author?
No. Overwhelmingly journal publishers and research communities have stated that that AI models cannot meet the responsibility and accountability criteria necessary for authorship. AI models are non-legal entities and as such cannot assert the absence of conflicts of interest nor manage copyright and licensing agreements. Generative AI models are also unable to be held accountable for the liability and answerability for the work and thus cannot be considered as an author.
For More Information:
AI and Publisher Guidelines This website provides general guidance for each of the major academic publishers and provides links to each publishers AI guidelines.
Can AI be used in writing and/or developing manuscripts?
Maybe. Individual journals and research disciplines have different requirements regarding the use of AI in the writing process. In general, it is more likely to be tolerated when used for grammar and stylistic editing but is less tolerated and must be used with extreme caution in content creation. Publishers are advising authors using AI tools for the writing of manuscripts, production of images or graphical elements, or in the collection and analysis of data to be transparent in disclosing this use. Descriptions of the tool(s) used and how they were used should be included in the Materials and Methods (or similar section) of the paper. Authors must be responsible and accountable for the content produced by the AI tools they use and must ensure that AI-generated outputs are appropriate and accurate.
For More Information:
AI and Publisher Guidelines This website provides general guidance for each of the major academic publishers and provides links to each publishers AI guidelines.
The Use of AI and AI-assisted Technologies in Writing for Elsevier
The Use of Artificial Intelligence to Improve the Scientific Writing of Non-Native English Speakers This article reports on a study performed to test whether or not the careful application of AI tools could improve the writing and of non-native English speakers in the scientific disciplines.
Can AI be used in writing grant applications?
Maybe. The same concerns that exist when using AI for writing/developing manuscripts (see above) apply to writing grant applications. Such applications are to represent the original ideas of the applicant, their institution, and their partner researchers. Therefore generative AI tools should not be used to create the text for each application section but may be used (with extreme caution) to edit the author's original text for grammar or writing style. Applicants should read the application's directions thoroughly to determine the funding agency's rules for AI use and documentation.
Can AI be used in the peer review process?
No. The National Institutes of Health (NIH) and other large funding agencies and scholarly publishers prohibit their peer reviewers from using natural language processors, large language models, or other generative AI for analyzing and formulating peer review critiques. Utilizing AI in the peer review process is to be considered a breach of confidentiality because such tools do not guarantee where submitted information and data are being saved, viewed or used in the future. Even using AI tools to help draft a critique or to assist with improving the grammar and syntax of a critique draft are considered breaches of confidentiality and are prohibited.
For More Information:
Using AI in Peer Review is a Breach of Confidentiality -- NIH Also The Use of Generative Artificial Intelligence Technologies is Prohibited for the NIH Peer Review Process
Using AI in Peer Reviews and Publishing -- Sage
AI-Assisted Peer Review This article, from the Humanities and Social Sciences Communications journal, reports on an experimental study done on using AI in the peer review process.
How should AI use be reported in my research?
Rigor and reproducibility standards are often established by specific journals and research disciplines. Transparent and complete reporting of methodology and materials used is crucial in promoting reproducibility and replicability. The Association of the Advancement of Artificial Intelligence has a helpful reproducibility checklist that can be found here.
For More Information:
AI and Publisher Guidelines This website provides general guidance for each of the major academic publishers and provides links to each publishers AI guidelines.