An integrated model of executive function, AI literacy, and Chinese L2 reading comprehension among ethnic minority adolescents in Hong Kong

Artificial intelligence (AI) is increasingly embedded in second language (L2) learning, yet the cognitive demands of AI-mediated environments, particularly for logographic languages such as Chinese, remain underexplored. This study investigated the interrelationships among executive function (EF), reading-related linguistic skills, AI literacy, and Chinese L2 reading comprehension in 155 Form 4 Southeast Asian ethnic minority students in Hong Kong. Using a quantitative design, we validated a three-factor EF model (working memory, cognitive flexibility, inhibitory control) and examined how these components, together with orthographic and morphological skills, relate to L2 reading outcomes. Structural equation modeling showed that all three EF components significantly predicted reading comprehension, underscoring the importance of executive control for processing a complex logographic script. AI literacy significantly supported both foundational reading skills and overall reading comprehension. A key finding of the integrated model is that cognitive flexibility significantly predicted AI literacy, which in turn significantly predicted reading comprehension, indicating an indirect pathway from EF to reading via AI literacy. This pattern highlights the cognitive underpinnings of digital engagement in L2 learning. The study extends the Reading Systems Framework by positioning AI literacy as an external scaffold that interacts with learners’ cognitive resources, and offers practical implications for designing cognitively supportive, AI-enhanced Chinese L2 instruction for ethnic minority learners.