The primary objective of this study was to develop a comprehensive model for fostering AI-driven data science literacy among students. Utilizing a meta-synthesis approach and a theory-building qualitative method, the research integrates precursors, processes, and outcomes into a unified framework. A systematic search was conducted using keywords such as "data science literacy," "artificial intelligence," and "machine intelligence" across prominent international databases (Wiley, ScienceDirect, Taylor & Francis, ProQuest, Scopus, and Emerald) as well as Persian databases (Magiran, Noormags, and SID). Following a rigorous screening process, 77 articles were selected for final analysis. Data were analyzed through grounded theory using open and axial coding. The findings revealed that the precursors of AI-driven data science literacy comprise three main themes: theoretical and strategic curriculum frameworks, teacher competencies, and student competencies (encompassing 59 open codes). The process of literacy development is structured into seven main themes supported by eight specific categories of digital tools and platforms. Additionally, the outcomes are classified into three core dimensions: academic-cognitive, metacognitive-affective, and social-ethical (spanning 15 open codes). The results indicate that this model empowers students for informed and responsible participation in a data-driven society. Ultimately, the study emphasizes that achieving robust data science literacy necessitates data-centric curricula and sustained professional development for teachers to transition from traditional roles to AI-augmented learning facilitators.