Marc Waldman, Lorrie Faith Cranor, and Avi Rubin, AT&T Labs-Research
Trust plays a central role in many aspects of computing, especially those related to network use. Whether downloading and installing software, buying a product from a web site, or just surfing the Web, an individual is faced with trust issues. Does this piece of software really do what it says it does? Will the company I make purchases from sell my private information to other companies? Is my ISP logging all of my network traffic? All of these questions are central to the trust issue. In this chapter we discuss the areas of trust related to distributed systems—computers that communicate over a network and share information.
Trust in peer-to-peer, collaborative, or distributed systems presents its own challenges. Some systems, like Publius, deliberately disguise the source of data; all of the systems use computations or files provided by far-flung individuals who would be difficult to contact if something goes wrong—much less to hold responsible for any damage done.
In the physical world, when we talk about how much we trust someone, we often consider that person’s reputation. We usually are willing to put great trust in someone whom we have personally observed to be highly capable and have a high level of integrity. In the absence of personal observation, the recommendation of a trusted friend can lead one to trust someone. When looking ...