An Empirical Study of Architectural Shift from Traditional to AI-Enabled Simulink Controllers

Learned models for industrial control are usually judged by aggregate accuracy, but accuracy at the component level does not guarantee safety once it is embedded in the system it is meant to serve. We study this gap on a behavior-cloning task: imitating an expert Nonlinear Model Predictive Control (NMPC) policy for load-following of a Pressurized Water Reactor (PWR), an industrial system with tight safety constraints. We propose a structured architecture encoding variables from each timescale into separate latent spaces, reflecting the physical decomposition of the system, before training a controller to imitate the expert on the product latent space. On long-horizon rollouts, separated embeddings improve both accuracy and feasibility compared with a shared-embedding baseline. Sensitivity analysis further shows that our model yields interpretable representations aligned with the system's physics. However, standalone deployment still leaves several percent of trajectories infeasible regardless of the architecture. Using our method to warmstart the NMPC optimizer rather than acting standalone, we recover full feasibility and near-optimal cost while still cutting computation time by $\sim$15% relative to the expert controller, and even more for abrupt operating changes.

AI safety is still largely framed as alignment: training models to follow human preferences, safety policies, and normative constraints. That framing has improved the behavior of modern language models, but aligned behavior does not by itself guarantee that a deployed agent can be stopped, overridden, or constrained once it operates in open-ended, interactive, and tool-using environments. A system may be safe in expectation and still fail to yield to explicit runtime authority under conflicting instructions, long-horizon execution, adversarial inputs, or risky tool use. This position paper argues that AI safety therefore requires controllability as a first-class objective. We define \emph{controllability} as the ability of an AI system to remain reliably interruptible, overridable, redirectable, and constrainable by explicit control signals at runtime while preserving ordinary utility when such signals are absent. To study this gap, we introduce \controlbench{}, a benchmark for evaluating controllability failures in high-risk agentic scenarios. Experiments with OpenClaw-based agents show that current alignment and guardrail mechanisms reduce risk, but often fail to provide persistent, authoritative, and enforceable runtime control. We therefore propose a control-centric architectural framework that highlights explicit control planes, runtime intervention pathways, persistent control states, and auditable decision interfaces as key design principles for future controllable AI systems.

Threat modeling for cyber-physical systems (CPS) remains a largely manual exercise. This project presents SMSI (System Model Security Inference), a hybrid neuro-symbolic pipeline that starts from a SysML architecture model and produces a prioritized list of NIST 800-53 security controls. The prototype has three main stages: a deterministic parser mapping system components to vulnerabilities via the NVD; a family of retrieval and classification models linking vulnerabilities to MITRE ATT&CK techniques; and a control recommender. We explore three approaches for CVE-to-ATT&CK mapping: a supervised classifier using fine-tuned SecureBERT+, retrieval-based dense encoders, and a zero-shot LLM approach using Gemma-4 26B. We validate the pipeline on a healthcare IoT gateway with nine software components. For the ATT&CK-to-NIST stage, pretrained SecureBERT achieves the highest control retrieval scores, demonstrating that dense embeddings provide a strong basis for automated control recommendation.