This paper proposes an intelligent connected vehicle (ICV) information security three-dimensional attack matrix model, which includes the following components: firstly, an algorithm for evaluating the security level of the electronic control unit (ECU) information is proposed to determine the X-axis content of the three-dimensional attack matrix model based on the security relevance level of the information; secondly, using the clustering method based on the Connected Vehicle Product Security Vulnerability Library (CAVD), the ICV attack nodes are identified to determine the Y-axis content of the model; and finally, based on the empirical database, common attack techniques for vehicles are identified to determine the Z-axis content of the model. The proposed model enables the visual display and analysis of ICV information security attack behaviors.