Files
Download Full Text (11.4 MB)
Document Type
Article
Description
Auditing GenAI literature search workflows, this article evaluates whether AI-assisted literature review retrieval is traceable, metadata-accurate, and reproducible under student-facing conditions. Across four systems and two retrieval postures, the study finds that natural-language prompting often produces unstable or unverifiable citation outputs, while Boolean translation improves auditability but not full reproducibility. The article offers a practical framework for accountable AI-assisted literature search, citation verification, DOI traceability, and transparent evidence retrieval.
Publication/Submission Date
3-24-2026
Keywords
AI tools; evidence accountability and synthesis; information retrieval; inquiry governance; large language models (LLMs); literature search; literature screening; metadata integrity; natural language processing (NLP); traceability; undergraduate and graduate research
Disciplines
Arts and Humanities | Social and Behavioral Sciences
Creative Commons License
This work is licensed under a Creative Commons Attribution 4.0 International License.
Recommended Citation
Leon, Cristo Ph.D. and Kudelka, Michelle, "Auditing GenAI Literature Search Workflows: A Replicable Protocol for Traceable, Accountable Retrieval in Student-Facing Inquiry" (2026). STEM for Success Resources. 131.
https://digitalcommons.njit.edu/stemresources/131
Comments
Acknowledgments
The authors acknowledge the administrative and technical support provided through institutional facilities and library access that enabled database searching, record management, and documentation of the audit workflow. All support, tool use, and contributions, including any GenAI assistance used during study design, data collection, analysis, interpretation, or manuscript preparation, are fully disclosed in the Disclosure of Support and Statement of Contributions (DSSC) (Document S2) included in the Supplementary Materials. During the preparation of this study, the authors used generative AI tools (as specified in the DSSC) to generate candidate citations, propose Scopus-style Boolean queries, and draft explanatory rationales as part of the audited workflows. The authors reviewed and edited all tool outputs and take full responsibility for the content of this publication.Conflicts of Interest
The authors declare no conflicts of interest.PDF Accesibility
Alternative text has been provided for all figures in the PDF version of this article to support accessible reading and compliance with digital accessibility practices.