Autonomous System Subversion Tactics: Prototypes and Recommended Countermeasures

Abstract

discusses the use of artificial intelligence and machine learning and their application to high-energy physics. As security researchers we are interested in their analysis of how deep neural networks are implemented with their physical systems and how their approaches to anomaly detection function such that we can assess applicability to our problem space.

Autonomous System Inference, Trojan, and Adversarial Reprogramming Attack and Defense (Final)

“…In light of this, our investigation extends beyond conventional cybersecurity parameters, diving into the intricate web of potential vulnerabilities woven into ML-based DTs and ACS in advanced reactor systems. A crafted cyber-physical testbed and preliminary ACS were devised to act as a mirror, reflecting potential configurations of advanced reactor control designs [7]. Moreover, this study is intertwined with a scrutinization of ML models, developed either through conventional, manually tuned methodologies or via automated means through AutoML, probing into their cyber-risk profiles within operational technology (OT) environments.…”