This study aims to identify and conceptualize the components of an artificial intelligence curriculum for primary education in the Islamic Republic of Iran. This applied qualitative study employed the classical grounded theory approach. Participants included 25 national experts in virtual education and information and communication technology curriculum selected through purposive sampling until theoretical saturation during the 2023–2024 academic year. Data were collected through in-depth semi-structured interviews and analyzed using open, axial, and selective coding procedures. Credibility and trustworthiness were ensured through participant validation and transferability techniques. Data analysis produced a comprehensive ten-dimensional curriculum model consisting of objectives, content, learning activities, tools, resources, scheduling, learning environment, evaluation, grouping, and underlying logic. Together, these dimensions form a coherent, future-oriented, and systematic framework for the design of an artificial intelligence curriculum in primary education and demonstrate an integrated developmental approach that extends beyond technical skills to cognitive, ethical, social, and cultural growth. The proposed model offers a scientifically grounded and contextually appropriate framework for designing, implementing, and evaluating artificial intelligence curricula in primary education and provides strategic guidance for preparing future generations to engage responsibly and effectively with intelligent technologies.