Neuro-Agentic Control: A Deep Learning-based LLM-Powered Agentic AI Framework for Controlling Security Controls
2026-07-10 • Artificial Intelligence
Artificial Intelligence
AI summaryⓘ
The authors address the problem of cyberattacks causing damage in industrial systems and the challenges of using large language models (LLMs) safely for automated control. They propose a new system that combines an LLM planner with a time-series model to check if suggested actions are safe before taking them. This method, tested on a water treatment dataset, outperformed other models by preventing more attacks and avoiding unsafe actions. Their approach shows how combining language models with physics-based checks can improve security in critical operations.
Operational TechnologyLarge Language Models (LLMs)Industrial IoTTime-Series Foundation ModelCounterfactual Physics InjectionSecure Water Treatment (SWaT) DatasetNeuro-Agentic ControlStochastic Attack ScenariosLSTMTemporal Convolutional Networks (TCN)
Authors
Saroj Gopali, Bipin Chhetri, Deepika Giri, Sima Siami-Namini, Akbar Siami Namin
Abstract
Cyberattacks on operational technology are increasingly causing costly downtime and physical damage, exposing the limitations of traditional rule-based monitoring in industrial IoT environments. While Large Language Models (LLMs) have strong semantic reasoning abilities to assist in decision support, their hallucinatory nature presents unacceptable safety liabilities for closed-loop control. This paper introduces a neuro-agentic control framework, a novel architecture that couples an LLM-based planner (i.e., such as Gemini 2.5 Flash-Lite) with a pre-trained Time-Series Foundation Model (TimesFM), to achieve physics-grounded autonomous defense. The paper introduces a ``Counterfactual Physics Injection'' mechanism that simulates the impact of LLM-proposed interventions within the numerical latent space of the foundation model before actuation, while allowing the system to reject hallucinatory or unsafe actions. Evaluated on an industrial dataset (e.g., the Secure Water Treatment (SWaT)) in the context of stochastic attack scenarios, the framework exhibited better performance compared to LSTM and TCN baselines. The Neuro-Agentic Loop prevented five breaches (33.3%) below the threshold versus LSTM (26.7%) and TCN (13.3%), with zero physically invalid (hallucinated) actions executed. These results demonstrate the efficacy of using foundation models as deterministic ``Sentinels'' to safeguard agentic AI in critical infrastructure.