Archives as wellspring, verifier and generator of knowledge: a vision for the age of AI
Abstract Archives and cultural heritage institutions increasingly offer services that combine large language models with retrieval-augmented generation, allowing users to put questions to their collections and receive answers grounded in archival records. Existing research has treated AI either as a tool for archival processes or as a source of knowledge production to be restrained, leaving open what archives should aim to become in the age of AI. Archival science, however, has already opened the way towards knowledge generation, at least at the level of theory, by regarding the archive not as a mere repository of records but as a site where the meanings of records are continually reconstructed. This paper proposes the knowledge-generating archive, which generates knowledge in response to users’ questions on the basis of its records, and explores the role of archives in the age of AI more actively by examining its strengths and limitations through a case study of letters from Korea’s Joseon dynasty. It proposes three principles for the knowledge-generating archive: generation and verification, interdependent internal principles of the archive, and the wellspring, an external principle through which the archive supplies human-made sources of assured provenance to a knowledge ecosystem increasingly filled with AI-generated text.
Authors
- Woojin Jung (ORCID: https://orcid.org/0000-0001-7806-597X)
- Joonmo Ahn
- Jongrok Lee
- Ilhoon Oh
- Chunghyun Kim
- Gyu-min Lee
Publication Details
- Journal
- Research Square
- Published
- 2026-11-13
- DOI
- https://doi.org/10.21203/rs.3.rs-11509514/v1
- Primary Topic
- Digital and Traditional Archives Management
- Type
- preprint