- Wednesday 18 February 2026
- 13:30 - 14:45 EST
- 18:30 - 19:45 BST
Metadata, Meaning, and Machines: Library Experiments with Generative AI
Session Description
As libraries increasingly explore generative AI tools to create, enrich, and reconcile metadata at scale, this panel brings together professionals from Hofstra University, West Virginia University, American University, Goldey-Beacom College, and Arizona State University—each collaborating with JSTOR Seeklight to advance metadata creation and enhancement using generative AI.
Panelists will discuss how their institutions are leveraging Seeklight to uncover hidden connections across collections, enhance description, and improve discoverability—while upholding ethical, privacy, and accessibility standards. These real-world examples illustrate how AI can complement human expertise in creating meaning from data, rather than simply automating tasks.
A moderator from ITHAKA will provide context for the Seeklight initiative and facilitate conversation around institutional motivations, lessons learned, and shared challenges such as data quality, transparency, handling personally identifiable information (PII), and ensuring informed consent in AI-generated contexts. Together, panelists and moderator will explore how libraries can responsibly implement generative AI while preserving the human judgment and interpretive depth that give metadata its meaning.
