Cybersecurity as an Application Domain for Multiagent Systems

The science of cybersecurity has recently been garnering much attention among researchers and practitioners dissat-isfied with the ad hoc nature of much of the existing work on cybersecurity. Cybersecurity offers a great opportunity for multiagent systems research. We motivate cybersecurity as an application area for multiagent systems with an em-phasis on normative multiagent systems. First, we describe ways in which multiagent systems could help advance our understanding of cybersecurity and provide a set of princi-ples that could serve as a foundation for a new science of cybersecurity. Second, we argue how paying close attention to the challenges of cybersecurity could expose the limita-tions of current research in multiagent systems, especially with respect to dealing with considerations of autonomy and interdependence.