In May 2026, we launched the first AI in Action cohort, providing guided support and peer learning opportunities to 20 academic libraries as they explored how to implement AI tools into their library’s operations. Since then, the cohort has made considerable progress: we have examined the range of AI use cases in academic libraries, reviewed the ethical considerations surrounding their use, and surveyed library staff across participating institutions about their perceptions of and engagement with AI. Additionally, at individual institutions, cohort participants began developing their local projects. As we prepare to launch the second AI in Action cohort in December 2026, we are taking stock of the current cohort’s progress and considering how its early lessons can inform the next iteration.

What we’ve learned

At the outset of the cohort, Ithaka S+R administered a baseline survey to understand how library staff were thinking about, using, and preparing for AI in their work. The survey received 867 responses, representing an estimated overall response rate of approximately 29 percent across participating institutions.

Many of the barriers to AI use that staff identified centered strategy and governance issues rather than access to these tools alone. Only 28 percent believed that their institution has provided sufficient guidance on how to responsibly use AI tools, and just 20 percent agreed that their institution has a clear strategy to guide the internal use of AI. Their views were comparatively more positive about access to AI tools and the technical infrastructure needed to support them. For example, 66 percent agreed to some extent that they had access to AI tools or features relevant to their work. These findings suggest that institutional readiness requires more than technical access; it also depends on clear policies and strategic direction.

Our findings also suggest that libraries may encounter considerable staff resistance and uncertainty as they implement AI projects, though staff may simultaneously recognize the potential benefits of incorporating new technology into routine work. Many respondents believed that learning to use AI tools could help them stay current with technologies affecting academic libraries (73 percent) and remain competitive for future career opportunities (63 percent). They also identified potential operational benefits for their libraries, including reducing routine or repetitive work, supporting data analysis, and streamlining workflows. Even so, concerns about using AI tools in library operations were extensive: 91 percent expressed concern about the quality of AI-generated outputs, 88 percent about data privacy and confidentiality, 87 percent about environmental impacts, and 77 percent about the ethical or moral implications of AI use.

What participating libraries are doing

The survey helped to surface on-the-ground, library-specific perceptions about AI, providing helpful context as the cohort participants begin to implement pilot projects within their institutions. These projects serve as the central purpose of the cohort, with each library introducing an AI-enabled tool or feature into its operations. Each has designated a project lead and assembled a core team of two to four staff members representing a range of library areas to carry out this work.

As they scope their project plans, teams are drawing on our knowledge of existing AI use cases in libraries, as well as resources addressing emerging ethical considerations. Each team is developing a distinct project designed to address a local need, though several common areas of focus have emerged across institutions. Several libraries are focusing on collections management, some on institutional repositories, and others on information discovery or operational documentation. When asked why they selected this area of focus, most teams reported that their projects could reduce repetitive work and the staff time required to complete it. They also saw opportunities to provide services that their libraries currently lack the staff capacity to support.

As the cohort continues, we look forward to supporting project teams as they put their pilot projects into action. One benefit of the cohort model is the opportunity for participants to learn from their peers. Future sessions will provide dedicated space for teams working in similar areas to connect with one another and with Ithaka S+R staff.

What’s next

Near the end of the cohort, we will administer a follow-up survey to examine how staff perceptions and experiences may have shifted following project implementation. In the meantime, it is clear from the baseline findings that libraries will need to communicate why they are introducing particular tools, how their use aligns with institutional priorities and the library’s mission, and what is expected of staff as they implement their pilot projects.

Indeed, a consistent theme across the survey findings and conversations with cohort members is that clear institutional strategies and guidance are critical to helping staff engage with AI tools responsibly and with confidence. These topics will receive greater emphasis in upcoming cohort sessions and will be a cornerstone of the second iteration of AI in Action. By ensuring that AI use is guided by clear institutional policies and strategic intent, its application in library operations may be both more effective and more positively received by staff.

At the conclusion of the cohort, we will publish a full report detailing its activities and the findings from the pre- and post-implementation surveys of library staff.

Interested in participating?

The second iteration of AI in Action will launch in December 2026. If your institution is interested in participating, please send an expression of interest to Tracy Bergstrom (tracy.bergstrom@ithaka.org) by November 15, 2026. We will then provide a more detailed prospectus including the timeline, staff capacity expectations, and cost.

Explore other open opportunities to join Ithaka S+R AI cohorts for higher education.