HathiTrust Secures $5M to Build AI Discovery Tools for 19M Books

HathiTrust has secured a $5 million grant from the Mellon Foundation to develop artificial intelligence infrastructure that will improve how researchers discover and analyze material in its collection of 19 million digitized volumes.

The five-year initiative, called Transparent Books: AI Infrastructure for Discovery and the Public Interest, represents a significant investment in academy-controlled AI tools designed specifically for scholarly and public-interest applications. HathiTrust will collaborate with Northeastern University Library and the nonprofit Authors Alliance on the project.

Building Better Metadata and Search

The project will deploy AI to enhance the quality of descriptive metadata and text within the HathiTrust corpus. Planned improvements include more accurate identification of subjects and languages, automated generation of summaries, and interactive search tools that allow users to explore both individual works and thematic collections.

Beyond discovery, the partners intend to provide verified, transparent sources of book text and images at scale for academic purposes. This infrastructure could support text and data mining, computational analysis, and the development of specialized language models trained on scholarly material.

Why it matters

This project addresses a critical tension in AI development: the need for high-quality training data versus concerns about commercial exploitation of copyrighted works. By creating noncommercial, academy-led infrastructure with transparent governance, HathiTrust is positioning libraries as an alternative to purely commercial AI platforms. The focus on sustainability and shared principles could establish a model for how cultural institutions balance technological capability with legal compliance and public trust.

Governance and Sustainability Focus

Alongside technical development, the project will establish shared principles covering legal frameworks, research integrity, accountability, and environmental impact. The partners emphasized that academic values and public interest will guide decisions about technology deployment.

The initiative will also examine financial sustainability models that allow services to continue supporting academic and noncommercial research without compromising their mission.

Mike Furlough, HathiTrust's Executive Director, said the project continues libraries' longstanding effort to make the scholarly record more accessible. "Libraries created HathiTrust to ensure that the results of mass digitization would address the unique needs of researchers and the academy," he stated.

HathiTrust noted the work will consider the interests of publishers, authors, scholars, and the public as it develops AI-powered discovery tools.

These details were first reported by Research Information.