This study looks at academic librarian use of AI software integrations and examines how this key information using population are connecting, or not, their usual software tools with artificial intelligence.
All seven library deans and directors surveyed report using generative AI plugins or integrations regularly or occasionally, compared with none of the 11 technical services, cataloging, and metadata librarians surveyed. Pressure from the top, but reluctance from the rank and file, is a characteristic of the emerging use pattern.
Based on responses from 66 academic librarians, the study reveals uneven adoption across professional roles. Overall, 25.76% report regular or occasional use of AI plugins or integrations, while 43.94% do not use generative AI at all. Among technical services, cataloging, and metadata respondents, 72.73% report no generative AI use.
Some of the report’s many findings are that:
About 26% of survey respondents use AI plug ins at least to some degree
Another 12% have abandoned plug in use after trying an AI plug in.
Library deans and directors are overwhelmingly more likely than librarians in other roles to use AI integrations and plug ins.
19.7% of those sampled use Google Workplace integrations.
Use of Plug Ins with meeting applications such as Zoom were a major use category, utilized by more than 27% of respondents.
Evidence of operational benefits is mixed. While 18.18% of all respondents report improvement in staff time required for routine work, 28.79% report worsening requirements for staff review or correction. Costs are rated worse by 19.70%, compared with 4.55% reporting improvement. Many respondents have not measured outcomes or remain unsure, underscoring the need for more systematic assessment.
Written responses describe practical applications ranging from administrative reporting and workflow scripts to developing instructional resources and checking how AI search tools represent library policies. Other librarians cite fabricated references, verification burdens, privacy concerns, and doubts about the value of additional AI spending.
The study provides comparisons by professional responsibility, institutional characteristics, and respondent demographics, helping library leaders and vendors identify promising applications, adoption barriers, and assessment priorities.