Open Multi-Agent Systems (MAS) have growing popularity and are predicted to have many applications in near future, as large scale distributed systems such as clouds become more widespread. However, A major practical limitation to open MAS is security.
The focus of the talk would be on prevention and detection of information leakage in agent interactions, defined by LCC (Lightweight Coordination Calculus). Language-based security will be briefly introduced and a Security Type System for LCC will be presented. As a case study, information leakage in a cloud computing scenario will be discussed.