Exploring the Future of Security with ChatGPT

This article demonstrates how to prototype LLM-assisted security workflows using OpenAI APIs with Elastic/Kibana alerts, showcasing use cases such as ELI5 explanations, summarization, and alert prioritization. Through sample alert data and iterative prompting, it highlights the importance of context (e.g., severity, risk score, timestamps) and field selection for better results, discusses tradeoffs around cost and accuracy, and outlines broader opportunities for LLMs in SOC workflows, including training, automation, and threat modeling.