The accelerated pace with which companies, governments and institutions embrace digital transformation is creating opportunities for economic prosperity, but also increases the threat landscape. Recent orchestrated cyber-attacks have revealed the unpredictability of the harm they can cause in our society, rendering the creation of new models that capture systemic risk more critical than ever. In this paper, we model the behaviour of one of the most prominent cyber-attacks: ransomware; in particular ransomware that propagates between organisations via the Internet. We draw concepts from epidemiological models of viral propagation to reason about policies that can reduce the systemic cyber-risk to the community. To achieve this, we present a compartment-based epidemiological model of predator-prey interactions, and run simulations to validate the importance of defensive controls that reduce the propagation of ransomware. Our model suggests that with specific defensive controls in place, other response policies may also become more effective. A prey policy to not pay the ransom may improve the ability of the victim population to recover; while information-sharing may reduce the number of organisations compromised, if certain conditions on the speed of threat-intelligence sharing practices are met. These results indicate the validity of the approach, which we believe could be extended to explore the impacts of a broad range of attacker and defender behaviours and characteristics of the digital environment on systemic risk.
