The capabilities of artificial intelligence (AI) tools are changing rapidly. It is hard to believe that just four years ago, ChatGPT had not yet been released. OpenAI launched it in November 2022, and today we have agentic AI tools that can do far more than most of us would have thought possible even a short time ago.
One significant consequence of this rapid growth is that the way people access information has shifted. Many individuals now turn to a generative AI tool instead of a traditional search engine.
AI Is Changing the Library Technology Landscape
Library vendors are taking notice. In his 2026 Library Systems Report, Marshall Breeding documents how vendors across the library landscape are incorporating AI into their products and services. Clarivate, EBSCO, and BiblioCommons are all moving forward with AI in some capacity.
Even systems we might not expect to be early adopters are moving forward quickly. For example, the National Library of Medicine’s PubMed now uses AI-powered semantic search to improve retrieval, and AI-driven machine learning is being applied to automated cataloging systems to process large volumes of data and reduce manual effort.
Keeping Up with Emerging AI Research Tools
Special libraries are navigating this landscape in their own ways, and special librarians face a real and pressing question: How do we best leverage AI tools that are increasingly embedded in vendor services, while also staying current with emerging AI research tools?
Platforms like Elicit, Perplexity, Semantic Scholar, and ResearchRabbit are becoming essential for special librarians to understand and to teach others to use critically. Check out the post I wrote about research-specific AI tools back in 2024. Somewhat surprisingly, many of the big players are still the same.
Policy, however, has not kept pace with adoption. The 2025 AALL State of the Profession report found that the majority of law libraries still do not have a formal policy governing the use of AI in legal research, a gap that will likely need to be addressed as these tools become further embedded in daily practice.
Evaluating How AI Tools Generate Answers
An early concern about using AI for research was its tendency to hallucinate or generate inaccurate and unsupported information. While hallucinations do still occur, more tools are being designed to be both extractive—meaning they identify and pull existing text directly from a source to answer a question—rather than purely generative. Special librarians should understand how the tools they use are balancing extractive vs. generative capabilities.
The Special Librarian’s Role in Responsible AI Use
What makes this moment particularly significant for special librarians is that the profession has always been defined by expertise in evaluating sources and connecting people with credible information. That expertise is still needed, and it is worth saying that special librarians can help colleagues, researchers, and patrons use AI well.
I encourage you to continue to stay current with AI developments and talk to vendors about how they are incorporating AI into their products. It will be important to understand how vendors handle the data that AI creates and collects.
Privacy and information ethics remain core library values. It will be important for librarians to be directly involved with vendors to ensure the information practices their organizations value are upheld.
0 Comments