The academic research journey no longer begins at the library. Over the last decade, dramatic declines in library website traffic confirmed what many suspected: patrons moved to the open web, using search engines, publisher platforms, citation trails, preprint servers, and research networks while completely bypassing library discovery entirely. Most libraries have yet to reckon with what that shift means for how they serve researchers, design services, or measure success.

And now the environment has shifted again.

Beginning in late 2024, AI-driven bots (some good, many bad) emerged as a significant and growing share of content access requests – not researchers, not students, not faculty. The explosive growth of large language models created industrial-scale demand for scholarly content, arriving on top of an access ecosystem libraries and publishers were still working to navigate. The result is a research environment where human behavior is increasingly hard to distinguish from machine behavior, and where the metrics librarians rely on to understand their communities may no longer reflect reality.

This session asks a question librarians need to sit with: in 2026, who, or what, is “the user?” Attendees will examine how the research journey has evolved, what AI-driven bot activity means for understanding patron behavior, collection decisions, and access infrastructure, and where agentic AI is taking us next.