Guidelines for the Use of Generative AI in Research
Effective Dates: 08/15/2024
Last Updated: 08/15/2024
Responsible University Administrator: Assistant Provost for Research
Approving Body: Office of the Provost
Policy Contact: Office of the Provost, The Citadel
Purpose and Scope
Generative Artificial Intelligence technology, or generative AI, offers opportunities to assist in student learning, enhance faculty scholarly activity, and contribute broadly to scientific discovery, while simultaneously presenting significant challenges related to information security, data privacy, authorship attribution, and the introduction of machine-based biases which can undermine academic, legal, and more broadly, ethical standards of integrity.
In light of these challenges, the Office of Research & Grants provides the following guidelines on the use of generative AI in the scholarly endeavors of faculty, staff, and students. These guidelines summarize the guidance provided by a variety of colleges and universities and are centered on four critical areas related to the use of AI in research. They are intended to supplement those guidelines provided in A Report on Generative AI prepared by The Citadel’s Provost-Appointed Generative AI Taskforce in Feb. 2024.
When considering the use of generative AI in scholarly work, keep in mind that faculty, staff, and students are ultimately responsible and accountable for the accuracy of any content produced by the use of such tools.
Guidelines
1. Understand policies and guidelines within disciplines, journals, and granting agencies
Policies and guidelines concerning the use of generative AI are evolving across disciplines, federal agencies, journals, and academic institutions. Violations of these policies could result in scholarly work being unpublishable or unfundable. It is the responsibility of faculty, staff, and students to be aware of applicable policies and guidelines governing the use of generative AI tools and to determine how appropriate they are for use in scholarly activity
2. Establish expectations with colleagues and students
Effective communication among faculty and staff, collaborators at other institutions, and subawardees is necessary to ensure all parties have the same expectations regarding how generative AI will be incorporated into projects. Faculty mentors should have regular conversations with their students about the use of generative AI in their projects, especially those challenges and restrictions related to its use within the context of the discipline.
3. Maintain confidentiality
Generative AI tools may expose personal and confidential data of research subjects to third parties, so researchers must avoid uploading this type information to avoid FERPA, HIPAA, and other data privacy violations. Moreover, faculty must avoid uploading unpublished research data into generative AI tools as the models may incorporate the data into responses to queries from other users, thereby making results publicly available before they are ready to be shared. The use of generative AI in the preparation of peer reviews for journal publications and grant applications are widely considered a breach of confidentiality.
4. Recognize ethical conflicts
Faculty, staff, and student researchers must consider whether the use of generative AI produces any ethical conflicts, including plagiarism, inaccuracies in content and citations, and other breaches of academic integrity. As with other methodological tools, those who use Generative AI in research must be transparent regarding its use and properly cite and acknowledge the use of the tool, including describing methods of its use, as appropriate.
Related Policies/Documents: A Report on Generative AI Prepared by The Citadel’s Provost-Appointed Generative AI Taskforce