1st International Workshop "Safety and Security in MultiAgent Systems"
New York, NY, USA, 20/07/2004
As agent technology begins to be deployed in applications where incorrect or inappropriate agent behaviour can have harmful effects ranging from embarrassment to financial cost to physical injury to humans, safety and security are two central issues.
Multiagent systems are being designed for ever more demanding and possibly hostile environments. For example, NASA has proposed missions where multiagent systems, working in space or on other planets, will need to do their own reasoning about safety issues that concern not only themselves but also that of their mission. Likewise, industry is interested in agent systems that can search for new supply opportunities and engage in (semi-)automated negotiations over new supply contracts. These systems should be able to securely negotiate such arrangements and decide which credentials can be requested and which credentials may be disclosed. Such systems may encounter environments that are only partially understood and where they must learn for themselves which aspects of their environment are safe and which are dangerous.
On one side agents should act "safely": they should be robust enough to to avoid damages due to unexpected events, or internal bugs in non critical parts of their architecture. On the other side they should act "securely": they should be robust enough to resist potentially harmful effects due to malicious behavior of other agents. This workshop aims to bring together researchers, academics, practitioners and students interested in ways of ensuring that agent behaviour is safe and secure.
While intelligent software agents and multiagent systems can provide many advantages over conventional software systems, they are also usually much more complex. Software agents often integrate such activities as deliberately planning to achieve their goals, dynamically reacting to unanticipated obstacles and opportunities, communicating with other agents to share information and coordinate actions, and even learning from and/or adapting to their environments. Because agents are often situated in dynamic environments, these activities must be carried out in a timely manner.
These aspects of agents make the process of verifying and validating the safety and security of these systems much more difficult than for conventional software systems. Hence, verifying reasonable behavior from the agent and from multi agent systems requires new and different techniques and perspectives. The process of verifying and validating the safety of these systems is often complicated by the systems being open to unknown agents. Little, if anything, might be known about these agents. We need to be able to structure the system to handle the situatin when these agents might have goals that are atagonistic to the overall system goals. Moreover, even is the system is not open to all agents, there may still be security concerns if different agents are under the administrative control of different entities (eg in multi-national space missions or orchestrated financial web services).
evento ospitante
evento colocato
funge da
evento origine per pubblicazione
evento contenitore per presentazione
evento ospitato per
evento colocato per
ArgMAS 2004 @ AAMAS 2004 • 
An RBAC Approach for Securing Access Control in a MAS Coordination Infrastructure (articolo in atti, 2004) — Alessandro Ricci, Mirko Viroli, Andrea Omicini
An RBAC Approach for Securing Access Control in a MAS Coordination Infrastructure (SASEMAS'04 @ AAMAS 2004, 20/07/2004) — Alessandro Ricci (Alessandro Ricci, Mirko Viroli, Andrea Omicini)