Security in agent-mediated negotiation frameworks

Abstract

Several researchers in agent systems have proposed various settings for agent-mediated applications. As is the case with most software systems, there are security implications for any design decisions. This paper analyzes three agent-mediated negotiation frameworks and investigates the implications of their design to the applications' security. We present an overview of the unique threats faced by these negotiation frameworks, discuss implications of the threats and evaluate countermeasures so far suggested.