{"id":"arxiv/2604.00227","url":"https://arxiv.org/abs/2604.00227","tags":["math.OC"],"score":0,"title":"Deception in Linear-Quadratic Control","domain":"math.OC","source":"arxiv","authors":["Yerin Kim","Haosheng Zhou","Alexander Benvenuti","Ruimeng Hu","Matthew Hale"],"reviews":0,"subject":"math.OC","summary":"Systems operating in adversarial environments may inadvertently leak sensitive information to adversaries. To address this challenge, we revisit the linear-quadratic control framework and introduce deception to actively mislead adversaries. Specifically, we consider a blue-team agent, observed by a red-team agent, that seeks to minimize a quadratic cost while introducing perturbations to its trajectories over time. These perturbations are designed to corrupt the red team's observations and, consequently, any downstream inferences, while remaining undetected by a red team using sequential…","category":"research","comments":0,"disputed":false,"division":"math","mentions":0,"discussed":0,"published":"2026-03-31","reactions":{}}