Deep learning is an important branch of artificial intelligence technology, which is widely used in the fields of voice, video and image processing and unmanned driving. The “double-edged sword” attribute of technology makes deep learning technology develop rapidly, but also causes people to worry about its risks and hidden dangers. This paper introduces the development status of voice forgery technology and voice print identification technology based on deep learning, and discusses the development path of voice print forensic identification in the era of deep learning.