Are Academic Libraries Ready to Teach Scholarly AI Discovery?

p strongAcademic-library leaders see some value in teaching scholarly AI tools, but adoption and perceived productivity gains remain modest./strong/p p Primary Research Group's study of 51 senior library leaders found that 15.38% of research-university respondents used ResearchRabbit. Women were more likely than men to use Elicit. Just over half of the sample&mdash51%&mdashviewed patron training in AI research tools as a modest priority, while 31.73% said the tools had produced no change in their own productivity./p p Training should therefore be evaluative, not promotional. ResearchRabbit, Elicit, Semantic Scholar and Connected Papers organize discovery differently and may expose users to incomplete coverage, opaque ranking or AI-generated summaries. Librarians can teach patrons to move from a promising lead to the underlying source and to compare results with conventional databases./p p A useful workshop can center on one research question, several tools and a transparent comparison of coverage, citations, export and verification. Libraries should also state whether they license a product, whether user activity is retained and which disciplines are well represented./p p Readiness does not require every librarian to master every platform. It requires enough shared expertise to help researchers choose a tool, recognize its limits and avoid mistaking a fluent summary for a complete literature review./p p  /p