S-AI-Cyber : A Symbolic Hormonal Architecture for Adaptive and Parsimonious Cybersecurity

This article introduces S-AI-Cyber, a novel cyber defense architecture based on symbolic hormonal orchestration and parsimonious agent activation. Inspired by endocrine signaling mechanisms, the system replaces traditional threshold-based or black-box AI models with a biologically grounded framework combining transparency, adaptability, and computational frugality.S-AI-Cyber relies on five interacting components : Gland Agents emit symbolic hormones in response to perceived threats; the HormonalEngine modulates signal propagation and decay; Specialized Agents perform detection, classification, response, and inhibition; a Cyber-MetaAgent coordinates agent selection based on hormonal profiles; and a MemoryAgent ensures contextual traceability. All decisions are made through symbolic propagation without predefined rule sets or opaque inference.The proposed system was validated through two simulated asymmetric cyberattack scenarios : a Slow Port Scan (low-frequency stealth reconnaissance) and a Fast DDoS Attack (volumetric flooding). Each scenario was evaluated over multiple time steps, tracking hormonal levels, agent activation, and response strategies. The results demonstrate that S-AI-Cyber activates agents only when necessary, with distinct hormonal signatures for low- and high-risk threats. The system displays both reactive efficiency and symbolic explainability, while ensuring minimal resource usage during benign phases.The article provides full pseudocode, Python implementation excerpts, and symbolic execution traces for reproducibility. The findings support the viability of symbolic hormonal modulation as a foundation for next-generation, explainable, and context-aware cybersecurity systems, particularly in edge, IoT, and resource-constrained environments.