Leaky Birds: Exploiting Mobile Application Traffic for Surveillance.

Over the last decade, mobile devices and mobile applications have become pervasive in their usage. Although many privacy risks associated with mobile applications have been investigated, prior work mainly focuses on the collection of user information by application developers and advertisers. Inspired by the Snowden revelations, we study the ways mobile applications enable mass surveillance by sending unique identifiers over unencrypted connections. Applying passive network fingerprinting, we show how a passive network adversary can improve his ability to target mobile users' traffic. Our results are based on a large-scale automated study of mobile application network traffic. The framework we developed for this study downloads and runs mobile applications, captures their network traffic and automatically detects identifiers that are sent in the clear. Our findings show that a global adversary can link 57% of a user's unencrypted mobile traffic. Evaluating two countermeasures available to privacy aware mobile users, we find their effectiveness to be very limited against identifier leakage.