AI-enabled assessment in higher education: Institutional culture and student meta-skill readiness across university tiers

Universities are under growing pressure to integrate artificial intelligence (AI) while fostering meta-skills such as adaptability, collaboration, creativity and strategic thinking. However, limited evidence explains how institutional culture and AI-based assessment practices are associated with students’ meta-skill readiness, and whether governance and ethics readiness strengthens these relationships across university tiers. This study used a cross-sectional survey of 1280 undergraduate students from 12 Chinese higher education institutions representing national key, provincial public and local/private tiers. Validated scales measured institutional culture, AI-based assessment practices and meta-skills, and the data were analysed using confirmatory factor analysis, covariance-based structural equation modelling, latent moderated structural modelling and multi-group SEM. The measurement model showed excellent fit. Institutional culture was positively associated with AI-based assessment practices (β = .49, R 2 = .24) and meta-skills (β = .21), while AI-based assessment practices were strongly associated with meta-skills (β = .58, R 2 = .55). Governance and ethics readiness significantly strengthened the AI-based assessment–meta-skills relationship, and cross-tier comparisons showed a stronger culture-to-practice pathway in national key universities. These findings position AI-enabled assessment as a culturally mediated dimension of academic management rather than a purely technical add-on.