Does Knowing What You Know Matter? AI Literacy Metaknowledge and Trust in AI

Human-AI collaboration is becoming central to the future of work, yet trust in AI remains a persistent barrier to effective adoption. Building on AI literacy and metaknowledge research, this study introduces AI literacy metaknowledge, that is, knowledge about one’s own AI literacy, and examines its role in trust in AI. Using a combined survey and skill-assessment study, we find that AI literacy metaknowledge is positively associated with trust in AI. We further show that objective AI literacy is positively associated with AI literacy metaknowledge and that AI literacy metaknowledge fully mediates the association between objective AI literacy and trust in AI. Contrary to expectations, subjective AI literacy is also positively associated with AI literacy metaknowledge. Our findings broaden metaknowledge research beyond overconfidence and identify AI literacy metaknowledge as a novel correlate of trust in AI. These insights are especially relevant for training, reskilling, and the design of AI-mediated work arrangements.