Social features are affecting different application areas, from traditional online social networks to economical networks, as well as the study of epidemical behaviors of pandemics and diseases. From the information point of view, also the Internet paradigm has been revised according to social behaviors of nodes, showing intrinsic properties that have allowed the definition of the Social Internet of Things paradigm. Furthermore, considering vehicular networks, social features have been spreading their influence, and defined the concept of Vehicular Social Networks, where mobile nodes show predictable social behavior. In this context, all the networking concepts should be updated accordingly, taking into account node social aspects and interactions. In this paper, we discuss the open challenge of how we should define a trust vehicle, relying on social networking aspects. We introduce the concept of node trustworthiness in a vehicular environment, based on the reputation degree, computed according to real-time interactions and past data. A node can be defined as trusted if exhibiting an acceptable reputation degree, and being successfully able to transmit data packet in a given environment. The framework of a network architecture for the computation of node trustworthiness degree in a vehicular scenario is also presented.