AI assisted cyber crisis management exercise scenario generation
Δημιουργία σεναρίου ασκήσεων διαχείρισης κυβερνοεπιθέσεων με τη χρήση τεχνητής νοημοσύνης
View/
Keywords
Τεχνητή νοημοσύνη ; Κυβερνοασφάλεια ; Ασκήσεις κυβερνοασφάλειαςAbstract
Crisis Management exercises are a key tool in achieving the objectives of the cybersecurity strategy of all European Union member states. The increasing number of cyberattacks targeting critical infrastructures has created the need for a continuous training framework to assess the readiness of entities managing critical infrastructures to respond promptly and accurately. The frequency and content of cybersecurity exercises must be able to simulate, as closely as possible, threats that could affect the actual operations of the targeted organisations.
Furthermore, cybersecurity exercises are an important tool for assessing the preparedness of participating entities and identifying weaknesses and vulnerabilities in their systems. By simulating security incidents, the opportunity is provided to address security incidents in conditions similar to real-life situations, applying relevant security measures and emergency plans already in place. This enables entities to make the necessary improvements and updates. Furthermore, these exercises promote information and knowledge exchange, enhance collaboration between participating entities, and strengthen a culture of cooperation, ultimately raising the level of cybersecurity for a country or organisation.
The purpose of this doctoral thesis is to introduce a novel methodology that automates and accelerates the generation of dynamic content for crisis management cyber exercises using machine learning and artificial intelligence techniques. This methodology will allow organisers of such exercises, with limited technical knowledge, to dynamically produce realistic scenarios by providing only minimal, high-level unstructured information such as: similar known attack incidents, articles, attack techniques or tools, and the type of critical infrastructure involved.
A trained artificial intelligence agent will be responsible for selecting the appropriate category of threats, attackers, and techniques, simulating a variety of attack paths that could be implemented based on the sector/category of the critical infrastructure under consideration. Ultimately, a structured technical exercise scenario will be produced, ready for execution, taking into account relevant current and future trends affecting that particular type of critical infrastructure.
Department
Σχολή Τεχνολογιών Πληροφορικής και Επικοινωνιών. Τμήμα ΠληροφορικήςNumber of pages
347Language
EnglishCollections
Except where otherwise noted, this item's license is described as
Αναφορά Δημιουργού-Μη Εμπορική Χρήση-Όχι Παράγωγα Έργα 3.0 Ελλάδα
Αναφορά Δημιουργού-Μη Εμπορική Χρήση-Όχι Παράγωγα Έργα 3.0 Ελλάδα