Trust in Computer Systems and the Cloud

CHAPTER 4Defining Trust in Computing

As we move away from trust and human relationships and focus more on computer systems, we should re-examine our definition of trust and compare it with some of the definitions we find that are more relevant to a non-human context.

A Survey of Trust Definitions in Computer Systems

In their work on distributed trust, Adbul-Rahman and Hailes extend some of the simpler models in the literature into the world of distributed trust, Abdul-Rahman and Hailes¹ use Gambetta's definition of trust, which we analysed in Chapter 1, “Why Trust?”; they note the importance of these three points in any conversation around trust. Though they state them slightly differently, using the language we have adopted, they note the importance of:

1. The subjective nature of trust
2. The inability of the trustor to monitor all actions that might affect trust
3. That levels of trust depend on how the trustor's actions are affected by the actions of the trustee

We can probably accept point 1, whilst being aware that subjectivity is not necessarily a problem if the scope of the trust is known and managed; and we have already noted the difficulties presented by point 2. The third point is either a truism because we accept that the trust relationship to the trustee is affected by the trustor's perception of the trustee's actions, or it is irrelevant because we believe that the trustor's actions are not directly related to the trust relationship in consideration. Despite these ...