Rethinking the Digital Divides: Sociodemographic Disparities in College Students’ Generative AI Use, Literacy, and Attitudes

• In Event: Digital Inequalities 3.0: New Problems and New Approaches

Abstract

Generative artificial intelligence (GenAI) tools are rapidly reshaping higher education, yet little empirical evidence exists on how student engagement with these technologies is stratified across sociodemographic groups. Drawing on original survey data collected in Fall 2024 from over 900 undergraduate students at a large public university, this study examines disparities in GenAI adoption, usage across domains, multidimensional literacy, informational environments, and attitudes. Guided by the digital divide framework and technology acceptance theory, the analysis reveals substantial variation by gender, race-ethnicity, nativity, language, and first-generation status. Importantly, women and racially minoritized students report lower technical and conceptual literacy but greater concern about equity, cultural representation, and environmental impacts. International and multilingual students report higher usage and paid access, suggesting compensatory patterns of engagement. The study contributes policy-relevant evidence supporting strategies that integrate GenAI literacy into curricula, faculty development, and governance frameworks to promote more equitable outcomes in higher education.

Author

• Ran Liu, University of Wisconsin-Madison