A new independent study conducted by Emerging Strategy on behalf of Clarivate examines how AI is impacting two core academic library workflows: metadata creation and cataloguing, and course reading list support.
Based on interviews with 11 library professionals across 8 institutions, the study explores how AI is being applied in practice through Alma Metadata Assistant and Leganto Syllabus Assistant. The study also features four institutional case studies including Brock University and Universidad Tecnológica de Bolívar, showing how libraries are using AI to tackle metadata debt, support more courses, and shift staff time from data entry to professional judgment.
Measured operational impact
The findings highlight measurable operational gains in areas where libraries face ongoing pressure. Participants reported reduced time spent on repetitive work, faster progression to review-ready outputs, and increased capacity to support high-volume workflows without additional staff. Reported results include a 30–60% reduction in manual effort, course list creation reduced from 15–45 minutes to 2–5 minutes, and metadata transcription reduced from hours to minutes for draft records.
Human oversight and responsible use
The study also reinforces that AI does not replace professional expertise. Interviewees emphasized that review, correction, and governance remain essential, with the strongest outcomes seen when AI is applied to high-friction workflows as a first-pass support layer.
Application across library workflows
Examples span both course readiness and metadata workflows. In course support, Leganto Syllabus Assistant reduces manual citation entry and accelerates reading list creation. In metadata workflows, Alma Metadata Assistant helps address backlogs, reduce transcription effort, and improve record completeness.
For library leaders, the report provides a practical view of where AI is already delivering value and what responsible adoption looks like in real workflows. The findings are based on institutional experiences and are intended to inform, rather than generalize, decision-making.
Explore the Academic AI Impact Study here.
