The underlying belief and knowledge models assumed by various kinds of authentication protocols have been studied for well over 10 years now. On the other hand, the related question of the generic trust assumptions, which underlie the settings where the protocols are run, has received less attention. Furthermore, the notion of trust, as it is typically defined, has more been based on the formal model used than the real user requirements posed by the application context and the actual people using the system. In this paper, we approach that problem from the users' point of view. We briefly describe what are the psychological bases on which typical people build their trust assumptions on, and consider how these are reflected in a typical e-commerce setting today. Given this background, we proceed to contemplate how the systems could be made more trustworthy by explicitly representing the trust assumptions and requirements, and how these digital expressions of trust could be instrume...