GenAI-Powered Autonomous Cyber Offense-Defense: An Explainable LLM Red-vs-Blue Simulation and Self-Defense Framework

Key Points

• The central aim is to develop a GenAI-powered framework for autonomous cyber defense and offense using large language models.
• Developed a closed-loop framework for cyber defense and offense using large language models.
• Implemented offensive and defensive strategies autonomously under various simulated attack scenarios.
• Evaluated framework's performance against traditional static defenses in dynamic cyber environments.
• The GenAI framework successfully adapted to 95% of simulated cyberattacks, outperforming static defenses.
• Achieved a 30% reduction in response time to threats compared to conventional methods (p<0.01).
• Demonstrated improved decision-making capabilities with an accuracy rate of 85% in simulated engagements.