How Alma and Primo help academic libraries adopt responsible AI within their library management system and discovery solution
AI is a sea change for academic libraries. Seemingly overnight, AI transformed how students search for information, how collections are discovered and how library teams manage growing demands and workloads.
The challenge is not whether to adopt AI, but how to do so responsibly without undermining trust, transparency or professional expertise.
Insights from the whitepaper Responsible AI for Academic Libraries: Igniting the power within existing workflows, while keeping librarians in control reveal how libraries worldwide are navigating this challenge. According to The Pulse of the Library 2025*, conducted by Clarivate, 67% of libraries globally are now exploring or implementing AI, yet most remain in the early stages of adoption. Rather than chasing experimental tools, librarians are prioritizing solutions that fit naturally into existing workflows and reinforce professional judgment.
Why embedded AI matters for libraries
One of the clearest themes emerging from librarian feedback is that AI delivers the greatest value when it is embedded within core systems, not introduced as a standalone experiment.
“The most effective AI tools are the ones that meet people where they already are. When AI is embedded in discovery, it helps users move forward with their research without changing how they engage with the library,” comments Michael Gonzalez, University Librarian, University of Technology Sydney (UTS)
The Alma library management system and its integrated Primo discovery solution offer solid examples of how to make this happen. Their AI capabilities are built directly into the workflows libraries and their users already rely on, making them low risk, high impact entry points for innovation. Instead of requiring infrastructure changes or workflow disruption, these tools support incremental progress, allowing libraries to move forward at their own pace while maintaining confidence and control.
Supporting student learning and discovery with Primo Research Assistant
For many academic libraries, discovery and student research support represent the most visible and immediate opportunities for AI. In fact, The Pulse of the Library 2025* reports that 64% of academic librarians cite supporting student learning and improving content discovery as their primary AI objectives.
To meet rising expectations for fast, credible academic research, Primo Research Assistant brings AI directly into the discovery experience students already trust. Designed specifically for library environments, it is grounded in the Ex Libris Central Discovery Index, ensuring results are academic, curated and transparent.
Students can ask natural language questions, receive AI generated summaries based on top ranked library results, and access clear citations with direct links to full text. Importantly, the complete list of search results remains visible, reinforcing academic rigor and encouraging verification rather than passive consumption.
As Sue Stevens, Head of Library Systems at Cardiff University, explains:
“Even though we’re using AI, we still want our students to be checking and verifying what the AI has done. Primo Research Assistant makes it fairly easy to do this.”
Improving metadata quality while keeping librarians in control
Behind the scenes, AI is also helping libraries manage one of their most complex and critical responsibilities: metadata. As collections continue to expand across formats, languages and sources, maintaining accurate and consistent metadata is essential for discoverability, but increasingly time consuming.
The Alma Metadata Assistant was developed in close collaboration with librarians to address this challenge. Rather than replacing professional expertise, it augments cataloging workflows by assisting with routine tasks such as generating subject headings, summaries and language tags.
As Mingyan Li, Metadata Librarian at the University of Illinois Chicago, explains: “With this powerful tool [Alma Metadata Assistant], we can improve our efficiency and workflow, and guarantee at the same time expert interpretation when needed to oversee and ensure the quality of the metadata.”
Start smart and build from there
For academic libraries navigating AI adoption, the message is clear: meaningful progress does not require disruptive change. By embedding responsibly developed academic AI into familiar systems, Alma and Primo enable libraries to start smart to build confidence, skills, and impact incrementally.
AI adoption works best when it feels familiar, transparent, and aligned with library values. With Alma and Primo, librarians remain firmly in control, using AI as a partner to enhance discovery, improve workflows, and support student success.
Download the whitepaper to explore how from the University of Technology Sydney (UTS) and Cardiff University have adopted AI all while improving discovery, streamlining workflows, and protecting the values that matter most. Learn how your peers are using Alma and Primo to move forward with confidence.
*Source: Pulse of the Library Survey 2025
