Decoding Adversarial Attack Patterns

The technological advancements witnessed by our society in recent decades have brought improvements in our quality of life, but they have also created a number of opportunities for attackers to cause harm. Before the Internet revolution, most crime and malicious activity generally required a victim and a perpetrator to come into physical contact, and this limited the reach that malicious parties had. Technology has removed the need for physical contact to perform many types of crime, and now attackers can reach victims anywhere in the world, as long as they are connected to the Internet. This has revolutionised the characteristics of crime and warfare, allowing operations that would not have been possible before.

In this document, we provide an overview of the malicious operations that are happening on the Internet today. We first provide a taxonomy of malicious activities based on the attacker’s motivations and capabilities, and then move on to the technological and human elements that adversaries require to run a successful operation. We then discuss a number of frameworks that have been proposed to model malicious operations. Since adversarial behaviours are not a purely technical topic, we draw from research in a number of fields (computer science, criminology, war studies). While doing this, we discuss how these frameworks can be used by researchers and practitioners to develop effective mitigations against malicious online operations.