Challenges for Dynamic Analysis of iOS Applications

Part 2: Malware DetectionInternational audienceRecent research indicates that mobile platforms, such as Android and Apple’s iOS increasingly face the threat of malware. These threats range from spyware that steals privacy sensitive information, such as location data or address book contents to malware that tries to collect ransom from users by locking the device and therefore rendering the device useless. Therefore, powerful analysis techniques and tools are necessary to quickly provide an analyst with the necessary information about an application to assess whether this application contains potentially malicious functionality.In this work, we focus on the challenges and open problems that have to be overcome to create dynamic analysis solutions for iOS applications. Additionally, we present two proof-of-concept implementations tackling two of these challenges. First, we present a basic dynamic analysis approach for iOS applications demonstrating the feasibility of dynamic analysis on iOS. Second, addressing the challenge that iOS applications are almost always user interface driven, we also present an approach to automatically exercise an application’s user interface. The necessity of exercising application user interfaces is demonstrated by the difference in code coverage that we achieve with (60%) and without (16%) such techniques. Therefore, this work is a first step towards comprehensive dynamic analysis for iOS applications