Title:Longitudinal Compliance Analysis of AndroidApplications with Privacy Policies

Abstract:Contemporary mobile applications (apps) are designed to track, use, and share users data, often without their consent, which result in potential privacy and transparency issues. To investigate whether mobile apps are transparent about the collect information about users and apps comply with their privacy policies, we performed longitudinal analysis of the different versions of 268 Android applications comprising 5,240 app releases or versions between 2008 and 2016. We detect inconsistencies between apps' behaviors and stated use of data collection, to reveal compliance issues. We utilize machine learning techniques to classify the privacy policy text to identify the purported practices that collect and/or share users' personal information such as phone numbers and email addresses. We then uncover the actual data leaks of an app through static and dynamic analysis. Over time, our results show a steady increase in the overall number of apps' data collection practices that are undisclosed in the privacy policies. This is particularly troubling since privacy policy is the primary tool for describing the app's privacy protection practices. We find that newer versions of the apps are likely to be more non-compliant than their preceding versions. The discrepancies between the purported and actual data practices show that privacy policies are often incoherent with the apps' behaviors, thus defying the `notice and choice' principle when users install apps.