AI EdgeLabs Host Platform Security is an eBPF-powered, AI-driven endpoint security solution designed to provide real-time prevention, detection, and response across workloads, containers, GPU clusters, edge nodes, and critical infrastructure. The agent is implemented in Rust and operates with a low system footprint (typically under 5% CPU and 250 MB RAM). It performs kernel-level enforcement by intercepting and blocking malicious activity before it can execute or propagate, targeting zero-days, supply chain exploits, and APT-style attacks. Security logic runs entirely on-device, enabling autonomous, inline protection in milliseconds without reliance on cloud connectivity. This design supports deployment in sovereign, offline, and air-gapped environments where 100% data locality is required. Detection combines signature-based methods, machine learning, and self-adaptive AI. Visibility extends to process execution, system calls, network traffic, and file access, all mapped against the MITRE ATT&CK framework. The platform includes an LLM-based AI Security Assistant (AISA) that provides real-time alert interpretation and compliance-ready reporting to reduce analyst workload. It also generates AI-driven response playbooks tailored to each environment to automate incident response and reduce mean time to remediate (MTTR).