<![CDATA[Artificial intelligence (AI) literacy is often assumed to translate directly into learning benefits, yet emerging theory suggests that the quality of human-AI collaboration — not literacy alone may be the mechanism that converts competence into benefit. This gap remains largely unexamined among students in Indonesia's non-formal secondary equivalency programmes. This study develops and empirically tests a structural model examining the direct effect of AI Literacy on Learning Benefits and its indirect effect through Human-AI Collaboration among Paket B and Paket C students. Using a quantitative explanatory design with purposive sampling, data were collected from 48 students via three adapted Likert-scale instruments and analysed using Partial Least Squares Structural Equation Modelling (PLS-SEM) with bootstrapping. Results show that AI Literacy significantly predicted Human-AI Collaboration (β = .714, p < .001), and Human-AI Collaboration significantly predicted Learning Benefits (β = .760, p < .001), while the direct path from AI Literacy to Learning Benefits was non-significant (β = -.147, p = .507); the indirect effect was significant (β = .542, p < .001), indicating full mediation. Findings suggest AI-literacy curricula in non-formal education should be paired with explicit collaboration-oriented pedagogical support rather than technical literacy training alone.]]>
Copyrights © 2026
