AI applications are already changing the way libraries function and the role of librarians. Sidhu Haded They argue that while AI cannot simply replace librarians, libraries and librarians risk being left behind by users if they do not take advantage of new tools.
A 2024 study on AI in Indian libraries found that only 16.8% are using AI chatbots, 12.7% are using recommendation systems, and 26.4% are using smart shelving. These numbers reflect only basic AI adoption, especially when compared to more integrated applications in Western libraries.
While many in our industry are reluctant to engage with AI, we need to carefully consider how AI is already changing the way we work. Existing AI applications catalog, improve discovery, facilitate access, and facilitate data analysis. The question is not whether AI will be used to replace librarians, but how we use it carefully and intentionally. As libraries become leaders in AI adoption, they must become more resilient or face being left behind.
How are libraries using AI?
Libraries around the world are using AI for cataloging, metadata creation, user services, and resource recommendations. The Library of Congress is researching machine learning in several projects. Speech-to-Text Viewer allows you to search your audio collection via transcript. humans in the loop Staff and volunteers will be able to adjust metadata and modify automated output. Newspaper Navigator allows you to extract images and headlines from millions of historical newspaper pages to create open datasets.
The National Archives and Records Administration (NARA) is experimenting with AI to remove personal information in digitized government records and create initial descriptions of archives to improve privacy and efficiency. AI also assists with cataloging and metadata management. Machine learning can classify and tag large datasets. For example, a ChatGPT-based Dewey classifier can assign Dewey Decimal Classification (DDC) numbers based on title and subject. This has been a useful tool since OCLC discontinued its Classify service. In my own library, the tool classified most materials correctly, struggled with interdisciplinary content, but overall it was a time saver for large collections.
Chatbots and virtual assistants provide 24/7 support. The University of Oklahoma's chatbot Bizzy (Figure 1) answers Level I and II questions on the Reference Effort Assessment Data (READ) scale 24 hours a day, seven days a week.
Figure.1: Screenshot of Bizzy chatbot. sauce: University of Oklahoma Library – Facebook
AI also deepens accessibility and user engagement. The National Library of Sweden is developing a tool to train models on Swedish texts and transcribe radio broadcasts. The National Library Board of Singapore has adopted StoryGen, an AI tool that helps users reimagine folklore and classic stories.
Given its wide range of uses, these examples show that AI is becoming part of the user experience, rather than just a backend tool. Libraries are making their collections more engaging with personalized recommendations and interactive storytelling. They use AI tools such as text-to-speech, text-to-speech, and translation to improve accessibility. These features are especially useful for users with disabilities or for users who use educational materials in a language other than their native language.
How will the role of librarians evolve?
When AI can answer simple questions, organize metadata, and aid discovery, librarians can invest more time supporting research and communicating with the community. This change has raised concerns about the loss of traditional roles and even jobs. However, AI will not replace librarians. Rather, it complements our capabilities. Chatbots can handle simple queries and predictive analytics can help develop collections, but librarians are still needed to interpret the results, guide them, and include ethical oversight, and AI can't do that. Additionally, AI introduces new job roles where librarian knowledge is the focus.
- educator: Librarians have a unique opportunity to educate users not only about their searches but also to be skeptical of AI results. This includes explaining how algorithms work, how to design recommendation systems, and how to detect bias and misinformation.
- curator: In the age of AI, curation is not limited to the process of selecting books and magazines. This includes shaping the training data, selecting a collection to digitize and send to the AI system, and structuring the metadata in a machine-readable way. For example, when libraries use local language materials or community archives to train AI models, librarians must ensure that cultural context and language details are preserved.
- ethics guide: As AI systems become increasingly integrated into library services, librarians have a responsibility to become guardians of ethics. This means promoting transparency in algorithmic decision-making, user privacy, and opposition to surveillance-driven personalization models. In this regard, international professional bodies such as the International Federation of Library Associations (IFLA) and the Association of Research Libraries (ARL) have developed guidelines on privacy, human rights, accountability, and fair access. Such efforts demonstrate how librarians are expanding their functions as moral agents.
Strategies for building an AI-enabled staff culture
AI is changing libraries by improving librarians' jobs, rather than replacing them. However, from both a technological and ethical perspective, AI needs to be effectively deployed in the workplace where employees can actively engage with it. The main components are:
- Role clarity: Libraries need to define what AI will handle and which functions will continue to be human-driven. One such example is the Artificial Intelligence Library Services Innovative Conceptual Framework (AI-LSICF). This suggests that while AI can assist with recommendations and analysis, research guidance and collection oversight can be left to librarians.
- Continuous learning: Electronic Information for Libraries (EIFL) offers modules on AI literacy, ethics, and applications as modules tailored to local requirements. The program not only familiarizes librarians with AI tools, but also enables them to critically evaluate the tools' impact on people and collections.
- Collaborative design: Collaboration between libraries, technologists, legal experts, and users will help create AI services that prioritize inclusivity, transparency, and relevance.
- Ethical integration: When using AI in libraries, risks related to bias, transparency, and privacy must be considered. For example, when U.S. libraries implemented AI for diversity audits of their collections, staff raised concerns. They are concerned that these tools may oversimplify or misrepresent the needs of the community unless local librarians review and verify the findings.
Automated warehouse or expansion service?
In countries like India, many institutions lack the funding and technical capacity of Western libraries, and progress depends on affordable, scalable tools and strong public investment. While AI can enhance services, personalize learning, and strengthen the role of librarians as ethical guides, it also brings challenges such as reskilling, shifting professional boundaries, and new ethical risks. Libraries will need sustained organizational support to manage these transitions.
The future of libraries lies in a model where human expertise and artificial intelligence complement each other. Careful use of AI can improve services, make learning more personal, and emphasize librarians' role as ethics coaches. If we succeed, tomorrow's libraries will be less automated warehouses and more social experiences, where human wisdom and machine cognition will open doors and enrich the learning experience as well as the social purpose of libraries.
📨Like this post? Then sign up for our mailing list and get all new posts delivered straight to your inbox 📨
This article gives the views of the author and does not represent the position of the LSE Impact Blog or the London School of Economics. By leaving a comment, you agree to our comment policy.
Image credit: Coward on Shutterstock.
