With the rapid development of artificial intelligence, particularly generative AI, AI literacy is regarded as an essential competency for individuals and society. This research systematically organizes domestic and international literature and, through content analysis, addresses core issues in this field, including conceptual definitions, research approaches, developmental trajectories, core content, methodological theories and evaluation systems. The essence of artificial intelligence literacy has evolved into a multidimensional structure that encompasses knowledge, skills, mindset, attitudes and ethics. Three dominant research approaches are identified: the interdisciplinary-integration-oriented approach, the social-function-oriented approach, and the measurement-and-evaluation-oriented approach. Existing research has covered implementation strategies for multiple groups, the role of institutions such as libraries, and disparity studies. Both domestic and international competency frameworks are constructed around the “knowledge-skills-attitudes”, yet significant differences exist in group-specific and cultural perspectives. Current research faces several limitations, including inadequate validation of assessment tools, difficulty balancing generalizability and specificity, the knowledge-action gap stemming from reliance on self-report scales, lack of longitudinal and dynamic studies, andscarcity of cross-cultural applicability testing. Future studies should strengthen longitudinal research and measurement method innovation, prioritize developing culturally responsive frameworks, and address equity issues in literacy education.