<![CDATA[The massive integration of generative artificial intelligence (AI) in higher education, particularly in scientific writing, calls for a deeper examination of the cognitive foundations underlying students’ AI use. Although prior research has predominantly focused on AI adoption and attitudes, limited attention has been devoted to modeling the cognitive skills that meaningfully shape AI engagement. Addressing this gap, the present study develops and empirically tests a structural model of students’ AI use by investigating the roles of Computational Thinking (CT) and Deep Learning Skills (DLS).A quantitative correlational research design was employed, involving 273 undergraduate students from three teacher education programs. Data were collected using a structured self-administered questionnaire and analyzed through Partial Least Squares–Structural Equation Modeling (PLS-SEM) to evaluate both the measurement model (reliability and validity) and the structural relationships among constructs. The results indicate that both CT and DLS significantly predict students’ AI use in academic writing, with DLS demonstrating a stronger structural effect. The proposed model explains a moderate proportion of variance in AI utilization, suggesting that higher-order cognitive and learning competencies function as central determinants of effective and responsible AI engagement. These findings contribute theoretically by positioning AI use not merely as a technological adoption issue but as a cognitively grounded learning process. The study further implies that higher education curricula should systematically integrate CT and DLS development to ensure that AI serves as a cognitive augmentation tool that strengthens academic integrity and learning quality.]]>
Copyrights © 2026
