Does school-based AI education narrow readiness gaps? The role of prior agency-related learning

As artificial intelligence (AI) becomes part of everyday learning, preparing students to participate confidently and responsibly in AI-mediated environments is essential. Yet students enter AI classes with different prior agency-related learning experiences, indicated in this study by self-initiated AI learning before school-based instruction, which may shape how they benefit from formal curricula. Guided by Social Cognitive Theory, this study examined how school-based AI education and prior agency-related learning jointly influence changes in students’ AI readiness during the 2024–2025 academic year in Hong Kong secondary schools. AI readiness was examined across psychological readiness, including confidence, perceived AI knowledge, ethical awareness, social good orientation, and learning intention, and cognitive readiness, indicated by AI literacy test. Using a two-wave multigroup latent change score model, data from 752 junior secondary students (mean age = 14.06 years) were analyzed across three groups: high-agency learners (with self-initiated AI learning experience), curriculum-supported learners (with one year of school AI instruction but no self-initiated learning), and low-agency late learners (with no prior AI learning experience). All groups improved in both psychological and cognitive readiness after a year-long school-based instruction. However, while psychological gaps narrowed, cognitive gaps persisted, suggesting that school AI programs may foster psychological equity but not full cognitive parity. The findings show that students’ prior agency-related learning experiences shape how they benefit from school instruction. Inclusive AI education therefore requires not only curriculum access, but also sustained opportunities for self-initiated and meaningful engagement with AI.