Deception/Honesty Detection and (Mis)trust Building in Manipulable Multi-Agent Argumentation: An Insight

In manipulable multi-agent argumentation, each agent may transmit deceptive information to others for tactical motives. We contemplate epistemic states and their roles in deception/honesty detection and (mis)trust-building. We propose the use of intra-agent preferences for handling deception/honesty detection and inter-agent preferences for determining which agent(s) to believe in more. We illustrate how deception/honesty in an argumentation of an agent, if detected, may alter the agent’s perceived trustworthiness, and how that may affect agents’ judgement as to which arguments they should accept. A detailed comparison to an earlier study on deception detection highlights wider applicability of our approach.