NASIG 2026

his presentation introduces an experimental project that explores the use of artificial intelligence to support the creation of Library of Congress Subject Headings (LCSH) for library materials. As the library cataloging work faces the challenge of limited staff and original cataloging remains a time-consuming and advanced cataloging skills required process, libraries are increasingly interested in understanding whether AI tools can meaningfully assist subject analysis while maintaining the professional standards required for high-quality metadata. Our project investigates this matter by designing and testing an AI model trained on real cataloging records from UW–Madison Libraries and Miami University Libraries, focusing on East Asian language materials. This presentation reports methodology, workflow design, pilot results, challenges in data permissions, and possible implications for scalable AI-suggested cataloging.