Contextual Integrity and Privacy Enforcing Norms for Virtual Communities

Contextual Integrity has been proposed to define privacy in an unusual way. Most approaches take into account a sensitivity level or a “privacy circle”: the information is said to be private or public and to be constrained to a given group of agents, e.g. “my friends”. In the opposite, Contextual Integrity states that any information transmission can trigger a privacy violation depending on the context of the transmission. We use this theory to describe a framework that one can use in an open and decentralised virtual community to handle privacy in a socially enforced way. This paper describes a framework, in which we can formally describe privacy constraints, that are used to detect privacy violations according to the Contextual Integrity theory. This framework is part of an ongoing work aiming at applying social control to agents that handle the information, so that malicious agents are excluded from the system.