Artificial Intelligence Acceptance in Higher Education: Perceptions from Mexico, Spain, and Finland

This study describes university students’ acceptance of artificial intelligence (AI) in higher education across three institutional contexts in Mexico, Spain, and Finland. A quantitative descriptive-correlational design was used with a non-probabilistic convenience sample of 416 undergraduate students who participated in a structured virtual workshop on the academic use of AI tools. The study did not include random assignment, a control group, or a pretest–posttest comparison; therefore, the results are interpreted in descriptive and associative terms. Data were collected using the AIComplex instrument, which assesses perceived risk, performance expectancy, effort expectancy, facilitating conditions, perceived value, habit, perceived complexity, and AI acceptance in higher education. The instrument showed adequate overall internal consistency across the three contexts. Additional psychometric evidence was obtained through measurement invariance analysis by dimension and exploratory factor analysis. Descriptive statistics, independent samples t-tests with effect sizes, Pearson correlations, scatterplot matrices, and a correlation heat map were used to examine students’ perceptions and associations among the dimensions. The results showed generally favorable perceptions of AI, particularly in performance expectancy and perceived value. No statistically significant gender differences were found, and the effect sizes were trivial. The strongest observed associations with AI acceptance were found for perceived value, habit, performance expectancy, and perceived complexity. The exploratory factor analysis suggested partial empirical overlap among some dimensions, while the invariance analysis indicated that cross-context comparisons should be interpreted cautiously. The findings provide contextual evidence on AI acceptance in higher education, highlighting associations among cognitive, institutional, and experiential dimensions without implying causal or nationally representative conclusions.