Large Language Models (LLMs) are increasingly used to generate cultural insights, yet little is known about their cultural expertise and the extent to which their outputs replicate or correct for cultural stereotypes. This study compares country-level estimates of Individualism–Collectivism (I-C) derived from five state-of-the-art LLMs with a recently validated I-C index based on nationally representative surveys. The LLM-based estimates correlate highly with the survey-based index (up to r = .874) but also exhibit systematic biases: Western societies are consistently overestimated on individualism, while Confucian East Asian societies are underestimated. Statistically, the type of bias suggests group-specific baseline distortions rather than random noise, with smaller and older models showing weaker alignment. Moreover, the patterns bear the imprint of Hofstede’s legacy framework, indicating that LLMs absorb not only empirical information but also historical discourse. These findings demonstrate both the potential and the risks of using LLMs as cultural interpreters. Used responsibly – with calibration, human oversight, and awareness of their biases – LLMs can enrich cross-cultural relations and support global coordination. Used uncritically, they risk reinforcing outdated narratives that misdirect academic research and managerial decisions.