InfoSecPilot: Navigating the Complex Landscape of Information Security with an AI-Powered Knowledge Management Chatbot

Abstract

This research investigates the development and implementation of an AI-powered conversational agent utilizing large language models (LLMs) to enhance knowledge management capabilities for information security professionals. The study employed systematic prompt engineering methodologies and structured technology validation protocols to assess chatbot performance across multiple evaluation frameworks, including user satisfaction metrics, Cohen's Kappa inter-rater reliability analysis, and Confusion Matrix statistical validation. Empirical results demonstrate substantial concordance between AI-generated responses and subject matter expert assessments, with statistically significant accuracy rates and high user satisfaction scores. The findings establish the technical feasibility and practical utility of generative AI systems as sophisticated decision-support tools within information security practice domains. This investigation contributes empirical evidence supporting the integration of AI-assisted technologies in professional workflows, demonstrating measurable improvements in knowledge accessibility and evidence-based decision-making processes. The research represents a significant advancement in applying generative artificial intelligence to specialized professional contexts, providing foundational insights for broader adoption of AI-enhanced knowledge management systems in information security practice.