by Mike McArthur, IALL Secretary; Assistant Director, Goodson Law Library, Duke University School of Law
Legal research has always been an adaptive practice. From stone tablets to print reporters to digital databases, each transition has reshaped how we access and organize information. Yet despite these changes in format, the assumption that resources exist in a mostly stable, retrievable form has been consistent. The rise of AI tools is beginning to challenge that assumption.
At its core, legal research depends on permanence. Cases, statutes, and all forms of scholarship are more than just information; they are part of an enduring record. These resources are published, revisited, verified, and preserved over time. Our use of citations reflects this stability. Beyond acknowledging authorship, they point to something fixed. They tell us not only who made a claim, but where that claim resides and how it can be located again. This stability is not incidental; it underpins the reliability of legal systems and the research that supports them.
AI-generated outputs operate on a different model. They are not retrieved from a discrete, stable source but are generated in response to a prompt. The result is an ever-shifting, constructed answer that may never appear again in precisely the same form. In this context, information begins to resemble a response rather than a record. It does not “exist” in a fixed location—instead, it is created for immediate consumption and often disappears just as quickly.
Even when AI systems do provide citations, those references are shaped by processes that do not rely on analysis of the cited resources in the way traditional research demands. Even when anchored to a specific document, the algorithms are not designed to quote or paraphrase, but to calculate an output based on patterns and probability.
At the same time, AI is influencing not only how information is accessed, but also what qualifies as a source. As more content is produced with limited transparency about authorship or editorial oversight, the distinction between curated scholarship and generated text becomes less clear. In turn, the value traditionally associated with publication begins to erode. If publication no longer reliably signals authority, accountability, or permanence, its role as a marker of trust weakens. Together, these shifts move research away from permanence and toward ephemerality.
As AI tools become more embedded in research workflows, there is a risk that young users in particular will grow increasingly comfortable with answers that are not tied to stable, citable sources. These outputs are quick, confident, and often persuasive. They can feel complete but are intentionally designed to produce plausible responses rather than to direct users back to verifiable material. Over time, this can diminish the instinct to seek out underlying sources. When an answer appears sufficient, the need to locate, read, cite, or preserve the original material may feel less pressing. What begins as a matter of efficiency can gradually reshape expectations about what research requires.
This is where the role of the law librarian becomes more critical. If AI shifts research toward disposable outputs, librarians need to remain grounded in systems built on permanence. We work with materials that can be traced, evaluated, and revisited. We preserve the connections between information and its origins, even as those connections become less visible in everyday tools.
In this sense, law librarians function as a harbinger of such profound change. We operate between a growing reliance on generated responses and the continued need for stable, authoritative records. Our role increasingly involves translating between these modes, helping users move from an answer back to a source, from something that appears convincing to something that can be confirmed. It is not simply about identifying errors, but sustaining a research culture that prioritizes traceability, accountability, and the integrity of the record.
AI is not just altering research methods. It is reshaping expectations about what counts as an answer and where that answer comes from. The focus is shifting from fixed, citable sources to flexible, generated responses. The need for permanence has become more critical than ever.
Maintaining a connection to sources that endure, those that can be verified, revisited, and relied upon, remains essential. As that connection becomes less apparent in everyday research practices, preserving it requires deliberate effort. Increasingly, that responsibility falls to us as law librarians, who must help ensure that research continues to point somewhere stable, even as the tools we use move in a different direction.
This Blog contains entries by members of the International Association of Law Libraries on issues germane to the Association’s areas of focus. Views expressed in an individual entry only represent the views of the author, and not those of the International Association of Law Libraries or the author’s employer.