The useful question is what changes for users, developers or buyers, and whether the announcement stays industry context or becomes something people can actually use.

The UC Berkeley Library is piloting AI tools to assist with cataloging tasks, such as generating descriptive metadata for historical collections. Haiqing Lin, head of technical services, developed an interface using Google’s Gemini AI models to translate and describe 2,000 Chinese film posters from the Paul Fonoroff collection. The AI-generated drafts are reviewed by librarians before being added to the catalog, demonstrating how AI can support, rather than replace, expert curation. Lin emphasizes that librarians’ expertise remains essential in guiding AI applications.

The Library’s digital initiatives team, led by Carolyn Caizzi, is evaluating AI’s role in academic libraries amid broader technological changes. The approach prioritizes responsible use, testing where AI tools can assist and where they fall short. Caizzi highlights the Library’s adaptability, noting that librarians’ expertise in information management positions them to navigate AI’s integration effectively. The team is focused on ensuring AI tools align with the Library’s mission and ethical standards.

Becky Miller, the natural resources librarian, used an AI-powered metadata tool, JSTOR Seeklight, to process decades-old agricultural extension publications. The tool generated searchable metadata for materials previously inaccessible to researchers, with 79 records added to the Digital Collections portal. Miller plans to expand digitization efforts, noting the potential for these documents to aid climate researchers and historians studying California’s agricultural history. The project underscores AI’s role in unlocking hidden collections.

At The Bancroft Library, Christina Velazquez Fidler employs AI to detect sensitive information in born-digital archives, such as Social Security numbers. AI-generated Python scripts streamline the process of isolating sensitive data, reducing manual effort and accelerating the release of archival materials. Meanwhile, the Scholarly Communication and Information Policy team, led by Rachael Samberg, negotiates licensing terms to ensure researchers can legally use AI for text and data analysis. The team has also supported projects like digitizing films for computational analysis, demonstrating AI’s broader impact on scholarship.