Beyond Trust: A Belief-Desire-Intention Model of Confidence in an Agent's Intentions

Trust plays a fundamental role in multi-agent systems in which tasks are delegated or agents must rely on others to perform actions that they themselves cannot do. The concept of trust may be generalised and considered as a level of confidence in one's predictions of another agent's future behaviour. This has applicability beyond that normally ascribed to trust: for instance, one may be confident that a particular agent's intentions are hostile, and that this will be borne out by particular behaviours. In this paper we present a cognitive model of trust in which the central component is a Belief-Desire-Intention model or 'theory of mind' of a person or agent that evolves over time.