US Patent 11792229 AI-driven defensive cybersecurity strategy analysis and recommendation system

A system and method for automated cybersecurity defensive strategy analysis that predicts the evolution of new cybersecurity attack strategies and makes recommendations for cybersecurity improvements to networked systems based on a cost/benefit analysis. The system and method use machine learning algorithms to run simulated attack and defense strategies against a model of the networked system created using a directed graph. Recommendations are generated based on an analysis of the simulation results against a variety of cost/benefit indicators.