The university of maryland ischool, with the maryland state archives and kennard african american cultural heritage center, will explore the potential for incorporating language-based generative…

The university of maryland ischool, with the maryland state archives and kennard african american cultural heritage center, will explore the potential for incorporating language-based generative artificial intelligence (AI) technologies into future library and archival services while ensuring trustworthiness and mitigating potential bias and harm. This two-year applied research project will develop a user-centered chatbot to enhance exploration and discovery within library and archival collections. Other activities will include community-driven AI design and evaluation, as well as studying how to make archival collections AI-ready, comparing the results of AI queries with those produced from non-AI methods. The approach will be tested and evaluated through focus groups to develop generalizable approaches for collections. Expected outcomes include improved access to collections, enhanced user interfaces, and guidance on AI use in libraries and archives. The project will strengthen the ability of libraries and archives to serve the public and research communities by unlocking new dimensions of access, discovery, and understanding within their collections.