DOI: 10.69554/zmqc4019 ISSN: 2047-1319

The AI-powered archivist: Harnessing generative artificial intelligence for streamlined archival description

Anna Gibson Hollow, Lindsay Cline, Mohsen Kardar, Olesya Komarnytska, Laurel Warkentin
Show PDF Cite
This paper describes the Generative AI and Description Project started by the University of Alberta Archives team in January 2024. Stemming from a curiosity to explore the ways in which generative artificial intelligence (GenAI) can be incorporated into daily workflows, the project sought to find ways of streamlining and saving time on some of the more laborious tasks of archival description. The team focused the implementation of GenAI tools on the creation of biographical sketches and administrative histories. Three GenAI tools were chosen, based on their open access and familiarity to most — ChatGPT, Perplexity AI and Microsoft’s CoPilot. Person and organisation-based entities were selected for the trials. As the project progressed, detailed processing notes were taken to track the accuracy of each tool used in the development of an appropriate and useful biographical sketch or administrative history. The results indicated a varying degree of inconsistency and fabrication (ie errors or inaccurate information) across the three tools, highlighting the need for staff to proofread and confirm the validity of sources; thus, not saving significant time in the end. The goal of this project was to provide other archives professionals with specific examples of practical applications of GenAI tools in archival description workflows. We hope that this project can inspire others to explore other ways in which GenAI can become a part of our professional practices. This article is also included in The Business & Management Collection, which can be accessed at https://hstalks.com/business/.

More from our Archive