["Journal of Computer Assisted Learning, Volume 42, Issue 5, October 2026. ", "\nABSTRACT\n\nBackground\nThe initiation of K‐12 artificial intelligence education has been led by numerous governments and institutions. However, most existing meta‐analyses have focused on the effectiveness of AI tools for learning support (AI in education; AIED), rather than AI education, and research on comprehensive evidence on how AI education improves learners' AI literacy remains limited.\n\n\nObjectives\nThis meta‐analysis aims to provide an additional understanding of the future direction of K‐12 AI education by synthesising and analysing existing interventions in AI education classes to enhance learners' AI literacy and holistically discussing their effectiveness.\n\n\nMethods\nA total of 16 source articles were searched and analysed from 8 databases, considering our inclusion and exclusion criteria, yielding 57 effect sizes. These effect sizes were used to examine the effectiveness of various K‐12 AI education classes in enhancing AI literacy.\n\n\nResults and Conclusion\nThe analysis showed a significant overall mean effect of AI education classes (g = 0.892, CI: [0.548, 1.236], p < 0.001) on enhancing AI literacy. Moreover, moderator analyses by publication type, year of publication, and school level of the participants did not reveal any significant patterns, indicating that K‐12 AI education classes exhibit consistently positive effects across different contexts. The findings underscore the importance of continued research on K‐12 AI education and offer a strong rationale for its ongoing integration by policymakers across various governments. To support discussion on this topic, future studies can reference this research and investigate the topic with more published articles.\n\n"]