Abstract: The new-generation wind power technology is rapidly evolving toward larger scale, greater flexibility, higher intelligence, and enhanced coordination. Traditional aerodynamic research methods are facing significant challenges in accuracy, efficiency, and adaptability. Artificial intelligence (AI), with its powerful capability in high-dimensional nonlinear mapping and data-driven modeling, is progressively reshaping the research landscape of wind power aerodynamics. This paper provides a comprehensive review of AI applications in several key areas, including blade aerodynamic performance prediction, aerodynamic shape optimization, aeroelastic modeling of the entire turbine, multi-physics coupling in offshore wind power, and wind farm flow under field modeling with cooperative control. The results indicate that AI offers notable advantages in improving computational efficiency, integrating multi-source heterogeneous data, constructing high-fidelity surrogate models, and enabling adaptive control strategies. These advances are driving a paradigm shift in wind power aerodynamics from traditional experience-based or numerically driven approaches toward a data-physics collaborative framework. Nevertheless, current AI applications still face several common bottlenecks. These include the scarcity of high-quality training data, insufficient generalization capability across varying operational conditions, the inherent trade-off between real-time performance and model complexity, and a lack of physical consistency constraints in complex flow scenarios. Future research should move beyond purely data-driven paradigms and shift toward physically interpretable and multi-field coupled frameworks. In particular, for complex systems such as offshore wind power and large-scale wind farm cluster coordination, it is essential to establish a new-generation multidisciplinary research paradigm characterized by "AI-physics synergy". Such a paradigm would provide theoretical support and a methodological foundation for the efficient, safe, and intelligent development of wind power systems. This review aims to serve as a reference for further in-depth studies or practical applications of AI technologies in wind power aerodynamics.