Answers are grounded in this paper and cite their sources.
A growing body of research has documented demographic disparities in generative AI adoption by age, education, income, and race/ethnicity, but whether adoption differs by immigration status remains unknown. Using data from the Pew Research Center’s American Trends Panel (N=9,151), this study applied Blinder-Oaxaca decomposition to estimate the nativity gap in first-level AI adoption (ChatGPT awareness, ChatGPT ever use) and second-level AI adoption (specific use cases of entertainment, learning, and work tasks), and to determine how much of any observed difference is attributable to group characteristics versus differences in how those characteristics relate to outcomes. Findings show that foreign-born adults had higher rates of ChatGPT awareness (77.0% vs. 69.4%) and ChatGPT use (43.4% vs. 32.7%) after adjusting for age and gender. Among ChatGPT users, foreign-born adults were significantly more likely to use ChatGPT for learning (34.9% vs. 22.5%, p < .001). Age, education, race/ethnicity, and digital access were the primary contributors to the explained gap, though the relative importance of each factor varied across outcomes. For first-level adoption (ChatGPT awareness), sociodemographic differences accounted for 63% of the 9.9 percentage-point gap, with a significant unexplained component remaining (3.6 pp, adjusted p = .013) after FDR correction. For second-level adoption (ChatGPT use for learning), race/ethnicity alone accounted for nearly 80% of the explained gap. Recent immigrants (0-10 years in the U.S.) showed the highest ChatGPT engagement at the second level of adoption, and Spanish-language respondents reported lower first-level adoption of awareness (58.9% vs. 70.7%, p < .001) after adjusting for age and gender. These findings indicate that immigrant disadvantage in technology adoption does not extend to ChatGPT, and that language access and purpose of use are additional dimensions of AI inequality beyond nativity alone.
No takes yet. Share an insight, caveat, or question.
Yoo et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: