Your Name and Title: Brandy Whitlock, Professor + Instruction Librarian
Library, School, or Organization Name: Anne Arundel Community College
Co-Presenter Name(s): Cameron Caswell, Associate Professor + Reference Librarian; Christine Vasica, Associate Professor + Access Services Librarian
Area of the World from Which You Will Present: Arnold, Maryland
Language in Which You Will Present: English
Target Audience(s): Librarians, educators
Short Session Description (one line): Join a conversation about the ethical and practical concerns around acknowledging generative AI use in academic and professional work.
Full Session Description (as long as you would like):
There’s a lot of anxiety about how and when to “cite” generative AI (GenAI) use and output. This anxiety often seems to involve confusion around citing GenAI output as relevant information or evidence in a paper or presentation versus acknowledging the use of GenAI applications in research and composition processes (e.g., outlining, summarizing, analyzing, creating, editing, etc.). Some of this angst about GenAI attribution stems from uncertainties and competing ideas about the purpose of citation. How should we be talking to students and colleagues about the ethical and practical considerations around referencing GenAI output in academic and professional contexts, given that rationales about the function of citation continue to evolve? How might we build consensus across our libraries and campuses about how to approach issues around GenAI attribution?
We now have sample formats from both APA and MLA for citing ChatGPT output, as well as guidelines for acknowledging the use of GenAI in some research and composition processes, but there’s very little consistency across style guides. Chicago Style, for example, currently recommends using a footnote to describe GenAI-created content and not to include a bibliographic citation, while AMA currently recommends including only an in-text citation, similar to a personal interview. By establishing conventions for referencing GenAI output, are these style guides implying that GenAI provides the kind of quality information that scholars and researchers would trust and choose to reference over other kinds of information? How can GenAI output be more credible and reliable than, say, Wikipedia content, when GenAI’s output is produced from large language learning models that have been trained on Wikipedia, along with other Web 2.0 content that we so often discourage people from using in academic and professional contexts in favor of sources that are generally more authoritative, reliable, stable, and transparent? Should GenAI output be included in its entirety when it serves as a data point for original research? Are there other reasons to provide attribution for GenAI output in academic and professional work? These are the kinds of practical and ethical questions we will take up in this discussion.
Websites / URLs Associated with Your Session:
Replies