Charting a third path: Fair use, open access, and university-built datasets for academic AI

Presented by Bernard Becker Medical Library, Center for Health & Science Communication at Becker Library and WashU Libraries, in recognition of International Open Access Week.

Researchers who want to apply AI and text-data-mining methods to large collections face a familiar bind: the materials they need are hard to reach, and the two paths on offer both carry costs. Commercial AI developers gather text at scale with little transparency, while publishers increasingly ask universities to license every computational use. This talk makes the case for a third path. It revisits why fair use is a built-in safeguard for research rather than a loophole, drawing on the mass-digitization decisions that settled the point, and situates that doctrine within the federal open-access and information policy now facing legal challenge. It then turns to Transparent Books, a Mellon-funded partnership among HathiTrust, Northeastern, and Authors Alliance that gives academic researchers a community-governed, noncommercial way to build their own datasets for AI work grounded in scholarly values.

Dave Hansen
Executive Director
Authors Alliance

Dave Hansen is the Executive Director of Authors Alliance. Prior to joining the Authors Alliance, Dave was an Associate University Librarian and Lead for Copyright & Information Policy at Duke University Libraries, where he was responsible for core research and collections support for faculty and students. Dave has testified before Congress, the U.S. Copyright Office, and the USPTO on copyright, AI, open access, and fair use issues affecting academic authors and universities, and has experience teaching and advising researchers, publishers, and librarians on copyright, fair use, and authors rights. He received his JD and MSLS from UNC Chapel Hill.