A decision flow for mapping AI agents to the NIST cybersecurity framework 2.0

In this paper, I propose a decision-flow approach to systematically align artificial intelligence (AI) agent architectures with the six functions of the NIST Cyber Security Framework (CSF) 2.0. By categorising agents into reactive, deliberative, hybrid, and learning types, the framework guides organisations in selecting the most suitable AI agent based on specific cybersecurity threat profiles, operational requirements, and NIST CSF 2.0 subcategories. Through a layered decision model, security practitioners can quickly determine whether near real-time responsiveness, strategic planning, adaptive learning, or a combination of these capabilities is paramount. Preliminary examples demonstrate how the decision flow aids in effective deployment, from initial risk assessment to ongoing governance. The paper concludes with recommendations for integrating this agent-based strategy into existing Security Operations Centres (SOCs) and suggestions for future empirical validation.