When AI Answers First, What Happens to Reference Work?

Abstract

This exploratory qualitative study examines how generative artificial intelligence is reshaping reference and information services from the perspective of library workers. In response to growing concerns about hallucinated citations and fabricated sources, the research explores how library staff describe and respond to AI-related situations in their everyday practice. The study's data consisted of 19 comments collected in April 2026 from researcher-initiated discussion threads on Reddit. These comments, representing library workers across academic, public, science, and archival settings, were analyzed using iterative qualitative coding and thematic analysis. The findings identified four major themes: reference work as verification and repair; workflow adaptation under AI conditions; trust, authority, and librarian mediation; and selective and context-dependent AI use. The study concludes that AI is not replacing librarians but is instead shifting their labor toward the time-consuming verification and correction of plausible but inaccurate information. This shift reinforces the continuing importance of professional judgment and human mediation in digital information environments where credibility is increasingly difficult to determine. The study provides an initial understanding of how AI-generated misinformation consumes institutional labor and highlights the evolving role of librarians as essential mediators who help users navigate environments where plausibility can no longer be mistaken for credibility.