The apps installed on a smartphone can reveal much information about a user, such as their medical conditions, sexual orientation, or religious beliefs. In addition, the presence or absence of particular apps on a smartphone can inform an adversary, who is intent on attacking the device. In this paper, we show that a passive eavesdropper can feasibly identify smartphone apps by fingerprinting the network traffic that they send. Although SSL/TLS hides the payload of packets, side-channel data, such as packet size and direction is still leaked from encrypted connections. We use machine learning techniques to identify smartphone apps from this side-channel data. In addition to merely fingerprinting and identifying smartphone apps, we investigate how app fingerprints change over time, across devices, and across different versions of apps. In addition, we introduce strategies that enable our app classification system to identify and mitigate the effect of ambiguous traffic, i.e., traffic in common among apps, such as advertisement traffic. We fully implemented a framework to fingerprint apps and ran a thorough set of experiments to assess its performance. We fingerprinted 110 of the most popular apps in the Google Play Store and were able to identify them six months later with up to 96% accuracy. Additionally, we show that app fingerprints persist to varying extents across devices and app versions.

Although there are over 1,900,000 third-party Android apps in the Google Play Store, little is understood about how their security and privacy characteristics, such as dangerous permission usage and the vulnerabilities they contain, have evolved over time. Our research is two-fold: we take quarterly snapshots of the Google Play Store over a two-year period to understand how permission usage by apps has changed; and we analyse 30,000 apps to understand how their security and privacy characteristics have changed over the same two-year period. Extrapolating our findings, we estimate that over 35,000 apps in the Google Play Store ask for additional dangerous permissions every three months. Our statistically significant observations suggest that free apps and popular apps are more likely to ask for additional dangerous permissions when they are updated. Worryingly, we discover that Android apps are not getting safer as they are updated. In many cases, app updates serve to increase the number of distinct vulnerabilities contained within apps, especially for popular apps. We conclude with recommendations to stakeholders for improving the security of the Android ecosystem.

The adage "there's an app for that" holds true in modern app stores. Indeed, app stores usually go further and provide multiple apps with very similar functionality; examples range from flashlight apps to alarm clocks. We call these functionally-similar apps. When searching for these apps, users are often presented with a vast array of choices, but no distinction is made in the user interface to highlight the relative privacy risks inherent in choosing one app over another. Yet the availability of many functionally-similar apps raises the question of whether some apps are significantly less invasive than others. In this paper, we take several steps toward answering this question. We begin by enumerating 2 500 groups of functionally-similar apps in the Google Play Store. Within groups of apps, we use static analysis to understand the real-world risks coming from apps with aggressive permission usage. By leveraging an established ranking system, and combining it with real-world data from over 28 000 Android devices, we quantify the improvements that can be made if users installed apps with privacy in mind. We observe that at least 25.6% of apps contain libraries that gratuitously exploit available permissions and find that 43.5% of apps could be swapped for comparable alternatives that require fewer permissions. Permissions saved may deliver important privacy and security improvements, including preventing access to the calendar (in 24% of cases), sending text messages (12%) and recording audio (8%). This is particularly important for apps which embed third-party libraries, since library code executes with the same permissions as the app itself.

We demonstrate SecuRank, a tool that can be employed by Android smartphone users to replace their currently installed apps with functionally-similar ones that require less sensitive access to their device. SecuRank works by using text mining on the app store description of apps to perform groupings by functionality. Once groups of functionally-similar apps are found, SecuRank uses contextual permission usage within groups to identify those apps that are less permission-hungry. Our demonstration will showcase both the Android app version of SecuRank and the web-based version. Participants will see the effectiveness of SecuRank as a tool for finding and replacing apps with less permission-hungry alternatives.

Automatic fingerprinting and identification of smartphone apps is becoming a very attractive data gathering technique for adversaries, network administrators, investigators and marketing agencies. In fact, the list of apps installed on a device can be used to identify vulnerable apps for an attacker to exploit, uncover a victim's use of sensitive apps, assist network planning, and aid marketing. However, app fingerprinting is complicated by the vast number of apps available for download, the wide range of devices they may be installed on, and the use of payload encryption protocols such as HTTPS/TLS. In this paper, we present a novel methodology and a framework implementing it, called AppScanner, for the automatic fingerprinting and real-time identification of Android apps from their encrypted network traffic. To build app fingerprints, we run apps automatically on a physical device to collect their network traces. We apply various processing strategies to these network traces before extracting the features that are used to train our supervised learning algorithms. Our fingerprint generation methodology is highly scalable and does not rely on inspecting packet payloads, thus our framework works even when HTTPS/TLS is employed. We built and deployed this lightweight framework and ran a thorough set of experiments to assess its performance. We automatically profiled 110 of the most popular apps in the Google Play Store and were later able to re-identify them with more than 99% accuracy.