Are Chinese Preschoolers Ready for the AI Era? Evidence from a Large-Scale National Study in China

Research Findings: This study addressed a measurement gap in early childhood AI education by developing and validating the Chinese Preschoolers’ AI Literacy Scale (CPALS). Grounded in the five-dimensional SIACC framework – Safety, Identity, Attitude, Cognition, and Capability – the 41-item scale was administered to 10,341 preschoolers and showed strong reliability and validity. Cluster analysis with a refined sample of 8,512 children identified three unevenly distributed profiles: a dominant “Competent” group (50.59%), a large “Sufficient” group (37.37%), and a smaller “Emerging” group (12.04%). The “Competent” group scored highly across dimensions, especially Cognition and Safety. The “Sufficient” group was balanced but weaker in Identity and Attitude. The “Emerging” group lagged overall, particularly in Identity and Capability, while showing higher Cognition and Attitude. Practice or Policy: These internally unbalanced profiles show that one-size-fits-all AI education is insufficient. The SIACC dimensions help teachers observe how children protect information, distinguish AI from people, question AI content, and maintain their agency. Profile results can guide responsive pedagogy: routines for the “Emerging” group, reflective comparison for the “Sufficient” group, and open-ended, critical activities for the “Competent” group. Teacher preparation should train teachers to scaffold agency during play and inquiry. Policy should frame early AI literacy around agency, reflection, and emerging metacognition.