Librarians as guides in the age of AI

As research and higher education increasingly adopt AI-powered technologies, the role of librarians is evolving – both in guiding patrons through the changing AI landscape and in incorporating new tools into their own workflows. 

To ensure a sustainable AI future, libraries need a clear vision for effective integration. This topic was the focus of discussion among librarians at the Library Futures Forum during the Association of College and Research Libraries (ACRL) conference in April 2025. Participants explored AI’s future impacts on teaching and learning, and the evolving role of librarians. 

Rewriting the rules of teaching and learning

A recent report shows that 28% of students surveyed already use AI daily or every couple of days, and a further 28% use it on a weekly basis (Librarian Futures Part 4, 2025). This trend only looks to set to continue as access and use of AI tools for academia grows. At the Library Futures Forum, Library participants were keen to explore the different ways AI is transforming teaching and learning for faculty and students at their institutions. 

A turning point for critical thinking

Forum participants noted that effective AI use requires its own kind of critical thinking: knowing how to prompt, interpret, and evaluate results. The way a user prompts AI to produce the required output requires a degree of AI literacy and information evaluation – in a way, it’s like knowing how to find the right resources in a database.

One of the key takeaways from the Forum was that critical thinking itself will adapt to meet the needs of a research landscape powered by GenAI. The adoption of AI in teaching, learning, and research is prompting librarians to reconsider the role of critical thinking in the process of acquiring, evaluating and creating knowledge.

Assessing assessment

As the conversation at the Library Futures Forum continued, librarians agreed that the rise of AI in higher education demands a reexamination of assessment models, to ensure they remain effective and relevant.

As student use of AI increases, assessments could evolve to incorporate its use or become more personalized to each individual student. In contrast, assessments could also become more focused on testing students’ critical thinking, creativity, and problem-solving skills by moving offline, through oral examinations, group work, presentations, and other formats which limit the involvement of AI.

Evolving assessment methods in response to a changing technological landscape is nothing new: as one librarian at the Forum noted, the arrival of the calculator changed the way math competency was assessed.

Tools for transformation

When asked about the features needed in AI tools, librarians indicated transparency, tool governance and human oversight as the most important aspects (Librarian attitudes towards AI 2024). In discussion at the Forum, librarians shared similar views on what requirements AI needs to meet to support research, discovery, and higher education most effectively.

Firstly, librarians asserted that AI tools for researchers need to be built on a high-quality base of trusted content from across the sciences and humanities. Without this solid foundation of quality data, librarians and their patrons can’t verify the outputs from AI tools.

Furthermore, AI tools need to be intuitive, interoperable and well-integrated into existing platforms, systems, and workflows, so that users can switch seamlessly between tools and tasks.

Librarians at the Forum also emphasized the importance of human oversight and a high level of transparency in AI tool development and use, to ensure proper ethical practices are adhered to.

The right tool for the job

There are a wide range of different AI tools that academics could leverage for various tasks in the classroom and the lab.

A participant noted that in research, AI tools could improve the discoverability of relevant content. Currently, researchers can struggle navigating hard-to-use databases that require specialized knowledge or specific query structures and search logic. With the right AI tools in place, content discovery can become quicker and easier, allowing researchers to spend more of their time analyzing results and designing experiments.

In a similar vein, different AI tools could be used to improve the process of conducting a scoping review, visualizing results and lab data, or enhancing a researcher’s specialist knowledge on a particular topic.

Librarians at the Forum agreed it was imperative that any changes to the research workflow were developed with a high level of human involvement. As one participant reflected, AI tools can help to improve a scoping review, but they cannot replace the human decision-making process.

With so many potential use cases for specialized AI technology, tool selection itself will become a crucial skill and is an area where librarians can play a critical role in supporting both students and faculty.

In a recent survey, 58% of students reported that they would feel confident that a library-approved tool aligns to their institution’s AI policy, therefore making them more likely to use it (Librarian Futures Part 4 2025). However, according to Lo’s 2024 survey, 70% of librarians believe they are not prepared to adopt GenAI tools in the next 12 months. This gap indicates a need for further professional development for librarians around AI tool evaluation – a topic which participants at the Library Futures Forum felt strongly about.

Future-ready librarianship

Conversation at the Library Futures Forum then shifted to focus on the changing role of the librarian in this new era. Participants agreed that as AI develops and permeates more deeply into the research ecosystem, library roles will evolve in step.

Some elements of the traditional librarian role will become increasingly supported by AI – for example, metadata management and collection development. In reference services, librarians are likely to interact with their patrons in new ways, like running workshops or providing resources focused on effective prompt engineering or information retrieval.

We might also see new specialist roles for librarians emerging around AI tool evaluation and integration, or a greater emphasis on roles which focus on engaging patrons outside the library. As one librarian suggested, perhaps the shift in roles will allow for more human-to-human interaction, which could make the librarian’s impact on their institution more visible.

Librarians at the Forum agreed that while early-career librarians may receive AI training as part of their formal education, well-established librarians already in role will likely need to seek out training and professional development opportunities proactively if they want to keep pace with the changing landscape.

Proactive, prepared, empowered

For years now, librarians have been working hard to develop AI literacy programs, construct evaluation frameworks, consider ethical implications, test AI tools, and so on. However, more needs to be done – and at a faster pace – as the technology itself continues to evolve, and the research community adapts itself in step. As one librarian at the Forum put it, “We need to be proactive instead of reactive.”

Ultimately, through conversation and exploration at the Forum, participants identified several ways to be proactive and prepare themselves (and their patrons) for a research future powered by AI:

Professional development: Librarians cannot provide guidance for their institutions without proper training themselves. With the right support and professional development opportunities, librarians will be well-positioned to guide their patrons through this new era of research and higher education.

AI tools designed for research: The research community needs AI which is tailored to meet their needs – based on trusted content, with verifiable outputs, transparency built-in, and a foundation of responsible, ethical AI principles.

Guardrails, governance, and guidance: Librarians can play an active role in developing AI policies for their institutions, leveraging their skills across information literacy, ethics, and privacy, for example.

With these elements in place, librarians can be proactive in establishing high levels of AI literacy within their institutions.

By building AI literacy across the research ecosystem, librarians can help safeguard its integrity and sustainability.

References

Technology from Sage. (2025). (rep.). Librarian Futures Part 4. Retrieved June 2025, from https://www.technologyfromsage.com/whitepapers/

Elsevier. (2024). (rep.). Librarian attitudes towards AI. Retrieved June 2025, from https://www.elsevier.com/insights/attitudes-toward-ai/librarians.

Lo, L. (2024). Evaluating AI literacy in Academic Libraries: A survey study with a focus on U.S. employees. College & Research Libraries, 85(5). https://doi.org/10.5860/crl.85.5.635