(The presentation is held in french)
The ongoing 4th industrial revolution already opened fantastic opportunities for Cyber crime, from the emergence of the Internet in the 80’s to the day-to-day reliance of both individuals and companies on cloud solutions and the Internet of Things of the 20’s. Individuals are mainly exposed to Cyber crime through the third party liability of companies: ereputation, Cyber harassment, release of Personal Information… Companies’ Cyber vulnerability management is key to enhance the protection level of individuals. Actions must be taken collectively and should be influenced by governments and public authorities. These actions must rely and evolve with the knowledge of the risk. Actuaries are key actors to provide companies and governments with the most up-to-date research and operational studies to risk modelling and protection solutions. First, this paper tackles the challenge of systemic Cyber risk modelling based on a Catastrophe approach and accumulation scenarios. Second, the modelling of the frequency is investigated with a strong focus on dependencies, for different kinds of economic player and attacks. More specifically, the study considers the limits of the Poisson and Negative Binomial distributions, autoregressive processes, discrete and continuous modelling based on self-exciting point processes of which Hawkes processes then a diffusion matrix based approach. Eventually, the paper highlights the main limit of Cyber odelling, lack of data, and reminds collective best practices of sharing data.