Gendered Patterns of Generative AI Adoption among University Students: A Cross-Cultural Survey of Chinese and International Users

Abstract

The fast spread of generative artificial intelligence has already become pervasive in university students' media practices, but there is still an inadequate understanding of the gender and culturally embedded factors underlying its adoption process. This paper examines the effect of gender and cultural background in isolation and combination on the frequency, diversity, and types of generative AI adoption by university students in China using a cross-cultural online survey of 412 respondents, including both Chinese students and students of other countries. As shown, statistically significant gender differences occur on all three dimensions – male respondents have more frequent AI use and greater use diversity, with a notable inclination toward AI coding help and entertainment purposes, whereas women are mainly engaged in AI academic writing and language translation. From a two-way analysis of variance (ANOVA), gender and cultural background interact to produce a significant small effect size (=.011), where the gender difference among international students is larger than that of Chinese students (mean difference=0.61 vs. 0.26). Regression analysis also supports the significance of both predictors with a negative interaction coefficient (β=-.15, p=.038), which implies that Chinese cultural orientation reduces gender differences. In all, the results provide a broader understanding of gender digital divide by shifting the discussion from the access stage to the usage stage, indicating that AI technology's institutionalization and the features of the domestic platform ecosystem might partly explain the gender divide in China.