The purpose of this study is to explain a model for the acceptance of artificial intelligence (AI) in the teaching-learning process, based on teachers' AI literacy, technological trust, and digital self-efficacy. In terms of its objective, this is an applied study; in terms of methodology, it employs a descriptive-correlational design with a quantitative approach. The statistical population consists of teachers employed in schools in Tehran, and the sample size—calculated using Cochran’s formula for a large population—is set at 384 individuals. Data were collected via a questionnaire, and structural equation modeling based on Partial Least Squares (PLS-SEM), implemented in SmartPLS software, was used to analyze the relationships between the variables. The study’s conceptual model examines the direct effects of AI literacy, technological trust, and digital self-efficacy on AI acceptance, as well as the influence of AI literacy on the constructs of technological trust and digital self-efficacy. Furthermore, the mediating roles of technological trust and digital self-efficacy in the relationship between AI literacy and AI acceptance are investigated. Theoretically, it is expected that AI literacy—by enhancing the ability to critically evaluate technology, fostering reasoned trust in intelligent systems, and promoting a sense of digital competence—will be associated with greater acceptance of AI in educational activities. This framework can serve as a foundation for designing professional development programs for teachers, enhancing AI literacy, and promoting the informed, effective, and responsible use of intelligent technologies in education.