In 2024 the AU Library formed an Experimentation Working Group was to explore the potential of AI to enhance workflows and operations within the AU Library. Four subgroups were created to focus on different areas: Business & Administrative Tools, Archives & Metadata, Data Analysis & Cleanup, and Ticketing & Support. Each subgroup experimented with various AI tools and documented their findings.
Overall, the findings suggest that AI tools can be valuable assistants for library tasks, but they require human oversight and intervention due to inconsistencies and inaccuracies in their output. The Business Tools subgroup found AI helpful for meeting facilitation, writing, and presentations. The Archives & Metadata subgroup found mixed results with AI for metadata generation and image description. The Data Analysis & Cleanup subgroup found AI useful for coding and data analysis but required significant human guidance. The Ticketing & Support subgroup was able to use AI for data cleanup and analysis but noted limitations in handling complex data formats.
The Working Group recommends continued experimentation with AI, focusing on identifying tasks best suited for AI assistance and developing ethical guidelines for AI use. Clear limits and human oversight are vital for successful integration in library workflows. The entire report is presented here and can also be viewed in the American University Research Archive.
As emerging artificial intelligence tools have the capability to significantly impact how research is conducted, how knowledge is constructed, and how libraries operate and assist their patrons, the University Library formed a working group to focus on experimentation with AI. The working group’s mission was to research possibilities for use of AI to improve internal workflows and enhance operations in the AU Library via active exploration and experimentation.
The group recruited interested library faculty and staff to identify projects so that all units in the library could become more comfortable with AI and learn together. Each subgroup was entrusted with finding tasks that AI could optimize, testing the application of AI to those projects, and making recommendations to library leadership about the feasibility of applying AI to library workflows.
This working group also supported the recommendations of the AI Exploratory Working Group to follow AI research, news, and tool developments in the future.
Expected deliverables of the AI Experimentation Working Group included: