Leveraging Large Language Models for Security Response Workflows

Learn how Microsoft Security Response Center leverages Large Language Models (LLMs) to streamline vulnerability management in this 40-minute Black Hat conference talk. Discover practical approaches to handling high-volume security reports through advanced LLM applications, focusing on three key use cases: extracting shareable vulnerability information, predicting report severity, and generating root cause analysis from crash dumps. Gain actionable insights into implementing fine-tuned LLM solutions for security response workflows, moving beyond basic chatbot applications to achieve meaningful operational improvements. Master techniques for utilizing organizational data to maintain consistent quality in security responses while effectively prioritizing critical vulnerabilities.