This report looks at how academic library directors and other leading officials are using and plan to use generative artificial intelligence in their work. The report looks the impact of AI on library productivity, and its use in a broad range of library functions, including but not limited to cataloging, data analysis, marketing, content creation, image tagging, research support and chatbot development. The report also presents detailed data on the extent of use of ChatGPT Gemini, CoPilot, Claude and other applications. The study also offers hard data on the extent of prompt sharing, and the degree to which internet searches start with an AI application rather than a traditional search engine.
Just a few of this 88-page report’s many findings are that:
-
16% of libraries use AI for personalized recommendations, with higher adoption in larger institutions.
-
AI in Image Management and Digitization: AI adoption for image tagging is at 12%, and for digitization tasks at 6%.
-
Chatbots and Predictive Maintenance: 16% of libraries use AI chatbots, particularly in larger institutions, while only 2% use AI for predictive equipment maintenance.1011
-
32% of libraries use AI for data analysis, with research universities leading at 40%. AI is also used by 36% of libraries for research support, especially in larger institutions.1213
-
AI Budget and Spending: Only 8% of libraries have a specific budget for AI, with research universities dominating spending. Most institutions allocate minimal to no official budgetary funding for AI.
The librarians surveyed foresee a broad range of impacts but especially in research and discovery, writing, information literacy and cataloging. AI is expected to have transformative effects in research and instruction services. Its use as a research tool was by far the most important and broadly used function to date.
Data in the report is reported in the aggregate for the entire sample and broke out by many institutional criteria such as enrollment size, tuition level, Carnegie class and other variables.