We propose a method to detect illegal arms trade based on investor knowledge. We focus on countries under arms embargo and identify events that suddenly increase or decrease conflict intensity. If a weapon-making company is trading illegally, an event that increases the demand for arms may increase stock prices. We find positive event returns for companies headquartered in countries with high corruption and low transparency in arms trade. We also suggest a method to detect potential embargo violations based on chains of reactions by individual stocks. The presumed violations positively correlate with the number of UN investigations and Internet stories.