Librarians And Media Collections Specialists

Frontier LLMs have crossed a quality threshold where they perform the reference interview function — clarifying patron needs, recommending sources, synthesizing information, and explaining research methodology — at parity with or above median professional librarian quality for the vast majority of patron queries. ChatGPT, Perplexity, Claude, and Google's AI Overviews have become the de facto first-stop research tool for undergraduate students, general public users, and working professionals. Reference transaction volumes at ARL libraries have declined every year since 2012, with post-2022 declines accelerating dramatically — some institutions reporting 40–60% drops in a single year.

OCLC's AI-assisted cataloging tools, ExLibris Alma's metadata management automation, and the Library of Congress Cataloging Lab's machine learning prototypes have operationalized automated record creation for standard trade monographs, serials, and digital resources at 85–95% accuracy. Institutions adopting these tools report 40–70% reductions in professional cataloging labor hours for copy cataloging workflows. The remaining manual cataloging work is concentrated in rare, archival, and non-English materials — a small fraction of most library collections and inaccessible to most cataloging librarians outside major research institutions and national libraries.