In today's increasingly digitalized world, ensuring information security for enterprises and organizations across various business types and industries is becoming an increasingly pressing issue, requiring advanced threat detection and response methods. This article presents a decision support system developed by the authors for identifying and classifying information security threats and automating the generation of response scenarios. The proposed architecture is a modular structure combining machine learning methods, expert systems, and explanatory artificial intelligence tools, which improves the accuracy of threat identification and risk assessment and enhances the confidence of specialists in the automated decisions generated by the system's intelligent component. The study included developing and testing the system using the UNSW-NB15 dataset, which contains network traffic information generated under laboratory conditions. The presented results demonstrate the potential for implementing the developed system in enterprise security services, helping to minimize damage from attacks on the information infrastructure of corporate and government information systems. An idea is proposed for further development of the system, taking into account the expansion of datasets, including new types of threats and response scenarios, and the introduction of online learning to adapt models to the dynamically changing security situation.