Detecting PII Leakage Using DPI and Machine Learning in an Enterprise Environment

With the emergence of new technologies, the threat of cyberattacks is becoming more prominent. Apart from direct attacks, hackers also try to steal personal data through indirect methods, i.e., social engineering. This threatens not only single individuals but enterprises as well. As a form of protection, it’s important to raise awareness about cybercrime and social engineering. However, sometimes humans can be tricked by smart social engineering attacks. This calls for the need to improve the defense through the implementation of other methods such as Data Loss Prevention (DLP) or new customized monitoring systems that would further mitigate the risk of personal information leakage. Therefore, this paper implements a network traffic monitoring system at the network edge that inspects packets and classifies personal information based on an ML model trained on synthetic data.