With the expansion of artificial intelligence (AI) technologies in higher education, profound transformations have occurred in roles, tools, and teaching-learning processes. In this context, action research education can also undergo meaningful change by leveraging the capabilities of AI. The present study aims to design a conceptual model to explain the requirements, processes, and outcomes of AI-based action research education in the higher education system. This study was conducted using a qualitative approach and a theory-building meta-synthesis method. A total of 269 academic sources—including articles, theses, and digital resources—were analyzed following a systematic search in reputable national and international academic databases. Data analysis was carried out using grounded theory methodology in three stages: open coding, axial coding, and selective coding. The findings indicated that the key antecedents of the model include the initial competencies of students and faculty members in the domains of knowledge, skills, and attitudes, as well as technological infrastructures and capacity building within universities. In the action research process, the use of generative AI tools plays a significant role in various stages of research, such as problem analysis, intervention design, and evaluation. Ultimately, the outcomes of the model are reflected in two areas: research competencies and the psychological dimensions of learning, including increased motivation, self-efficacy, enjoyment of learning, and reduced cognitive anxiety. The proposed model can serve as a guide for the design and implementation of research-oriented courses based on intelligent technologies and for enhancing the quality of learning in higher education.