U-M Workshop Explores Generative AI in Social Science Research, Calls for Clearer Guidelines

ANN ARBOR–On September 4, the University of Michigan’s Institute for Social Research hosted a workshop gathering over 50 experts from U-M, peer institutions, data archives, the Census Bureau, and the National Science Foundation (NSF) to discuss generative AI’s role in the social science research data lifecycle. Participants examined experimental AI applications, from code generation and survey design to literature summarization and metadata extraction. While attendees highlighted AI’s potential to streamline repetitive tasks and analyze large text datasets, they emphasized the necessity of human oversight and expressed concerns about reproducibility, “black-box” outputs, and inconsistent rules regarding sensitive data use. The group agreed that developing clear, secure, and transparent guidelines is critical for responsible integration of AI in data workflows. Several working groups have been formed to continue addressing these key issues identified during the workshop.

See this linked slide for more information on “Uses of AI through the Social Science Data Lifecycle.”