Jason Clark
Head, Research Optimization, Analytics, and Data Services
Montana State University
Doralyn Rossmann
Dean of Library
Montana State University
Artificial intelligence (AI) offers transformative potential for libraries and archives through enhanced metadata creation, discovery systems, and personalized recommendations. However, these technologies also present critical challenges around user privacy, information accuracy, and algorithmic bias stemming from training data. In response to these issues, the IMLS-funded Responsible AI in Libraries and Archives project has created the Viewfinder toolkit to facilitate values-driven AI in libraries and archives. In this briefing, project team members will discuss the tool’s development, which is based on new research involving a literature analysis, case studies, and workshops with library and archives administrators, practitioners, and users. We will then present a walk-through of the Viewfinder toolkit. When using the toolkit, practitioners reflect on the values and viewpoints of various stakeholders who are impacted by a particular AI implementation, and they then respond to a series of action-oriented prompts for responsible AI. The goal of the toolkit is to help library and archives practitioners reduce potential liabilities and harms, thus increasing our confidence in the ability of AI implementations to improve library services.