Evaluating the statistical realism of LLM-generated social science data Contributed by Yu Xie; received December 31, 2025; accepted April 8, 2026; reviewed by David B. Grusky and Michael Hout Significance Large language models (LLMs) enable the generation of data that could potentially be analyzed for social research. While the need for assessing the validity of such AI-generated data is widely recognized, we do not yet have a coherent framework for assessment.