AI-Assisted Cataloging Workflow in a Law Library: A Case Study with International Government Documents

This paper reports on a project conducted at the Jerome Hall Law Library of the Indiana University Maurer School of Law, in which a structured, human-in-the-loop workflow has been developed to integrate AI-generated MARC drafts, AI-assisted Library of Congress Subject Headings (LCSH) recommendation with validation, and professional librarian review together. Such a workflow was applied to perform original cataloging for a collection of international government documents from a variety of jurisdictions and languages. Both the quality of the AI-generated outputs and the practical efficiency of the workflow were quantitatively analyzed. The results suggest that, when supported by careful preparation of source materials such as complete and well-organized scans, thoughtful prompt design, and professional review, current AI tools are capable of producing original cataloging records of useful quality while requiring a relatively limited amount of professional review and correction time, even for materials that have traditionally been among the most difficult to process. The findings from this project provide new empirical evidence on the role AI can play in supporting original cataloging of complex international materials and help inform ongoing efforts that integrate AI tools into library technical services practice.