Information systems constitute an increasingly attractive target for attackers. Given the number and complexity of attacks, security teams need to focus their actions, in order to select the most appropriate security controls. Because of the threat posed by advanced multi-step attacks, it is difficult for security operators to fully cover all vulnerabilities when deploying countermeasures. In this PhD thesis, we build a complete framework for static and dynamic risk assessment including prior knowledge on the information system and dynamic events, proposing responses to prevent future attacks. First, we study how to remediate the potential attacks that can happen in a system, using logical attack graphs. We build a remediation methodology to prevent the most relevant attack paths extracted from a logical attack graph. In order to help an operator to choose between several remediation candidates, we rank them according to a cost of remediation combining operational and impact costs. Then, we study the dynamic attacks that can occur in a system. Attack graphs are not directly suited for dynamic risk assessment. Thus, we extend this mode to build dynamic risk assessment models to evaluate the attacks that are the most likely. The hybrid model is subdivided in two complementary models: (1) the first ones analysing ongoing attacks and provide the hosts' compromise probabilities, and (2) the second ones assessing the most likely future attacks. We study the sensitivity of their probabilistic parameters. Finally, we validate the accuracy and usage of both models in the domain of cybersecurity, by building them from a topological attack graph