Systematic Detection of Capability Leaks in Stock Android Smartphones Michael Grace, Yajin Zhou, Zhi Wang, Xuxian Jiang North Carolina State University 890 Oval Drive, Raleigh, NC 27695 {mcgrace, yajin zhou, zhi wang}@ncsu.edu
[email protected] Abstract mobile apps can be readily accessed and downloaded to run on smartphones from various app stores [2]. For exam- Recent years have witnessed a meteoric increase in the ple, it has been reported [22] that Google’s Android Mar- adoption of smartphones. To manage information and fea- ket already hosts 150,000 apps as of February, 2011 and the tures on such phones, Android provides a permission-based number of available apps has tripled in less than 9 months. security model that requires each application to explicitly Moreover, it is not only official smartphone platform ven- request permissions before it can be installed to run. In dors (e.g., Apple and Google) that are providing app stores this paper, we analyze eight popular Android smartphones that host hundreds of thousands of apps; third-party vendors and discover that the stock phone images do not properly (e.g., Amazon) are also competing in this market by provid- enforce the permission model. Several privileged permis- ing separate channels for mobile users to browse and install sions are unsafely exposed to other applications which do apps. not need to request them for the actual use. To identify Not surprisingly, mobile users are increasingly relying these leaked permissions or capabilities, we have developed on smartphones to store and handle personal data. Inside a tool called Woodpecker.