AI-REASON: AI-assisted Reliable and Explainable Analysis for Security OperatioNs

Purpose and goal

The project´s aim is to develop an AI-assisted, reliable and explainable decision support system for security operation centers (SOCs) in industrial environments. The goal is to support SOC analysts in prioritizing and analyzing large amounts of security data by reducing false alarms, improving contextual awareness, and enabling faster, more consistent and reliable decisions in cyber incidents.

Expected effects and result

The project will deliver a prototype of an AI-assisted decision support system for SOC environments, enabling faster security decisions, more reliable incident assessment, and explainable AI-based reasoning. It will generate context-rich incident reports, explainable outputs, and actionable recommendations. The expected effects include reduced alarm load, improved operational efficiency, and support for enhanced cybersecurity in industrial settings.

Planned approach and implementation

The project follows an iterative development process that begins with a requirements analysis in collaboration with SOC analysts to identify workflows, constraints, and data sources. Based on this, a system is designed that combines rule-based detection (IoC), machine learning, and LLM-based reasoning. Data is collected and annotated for model training, after which the solution is integrated into a SOC environment and continuously improved through testing and user feedback.