With the vigorous development of new energy technologies worldwide, electric vehicles and their supporting charging facilities are becoming increasingly popular. However, the spontaneous combustion accidents of electric vehicles and charging safety have attracted much attention from researchers. From the perspective of charging safety factors, this article summarizes the methods based on recent deep learning algorithms applied in charging fault early warning of electric vehicles and charging equipment. First, this article introduces the fault diagnosis process for electric vehicles and charging equipment based on deep learning algorithms. Then, we classify existing charging safety warning methods based on deep learning algorithms and compare them according to their characteristics and limits. Finally, this article looks forward to the opportunities and challenges in the follow-up research on fault warning of electric vehicles and charging equipment.