Wavefront sensing is an important part of adaptive optics system, which plays a key role in the fields of ground-based telescopes, laser transmission in atmosphere, wireless optical communication, laser nuclear fusion, and freeform surface optical measurement etc. Meanwhile, as a general advanced technology, deep learning has made revolutionary progress in many fields such as computer vision, natural language processing and so on. Using deep learning method to improve the wavefront sensor in adaptive optics system to achieve more accurate wavefront detection and adapt to more complex application scenarios is the development trend of adaptive optics, and also a new topic in the field of deep learning. This paper, introduces the application status of deep learning in adaptive optics wavefront sensing in detail. It also analyzes the research characteristics of different types of wavefront sensors, such as phase retrieval wavefront sensor and Shack-Hartmann wavefront sensor, and makes a summary at the end.