Abstract
The market for smartphone applications is steadily growing. Unfortunately, along with this growth, the number of malicious applications is increasing as well. To identify this malware, various automatic code-analysis tools have been developed. These tools are able to assess the risk associated with a specific app. However, informing users about these findings is often difficult. Currently, on Android, users decide about applications based on coarse-grained permission dialogs during installation. As these dialogs are quite abstract, many users do not read or understand them. Thus, to make the more detailed findings from security research accessible, new mechanisms for privacy communication need to be assessed. In our market experiment, we investigate how fine-grained data requests during runtime affect users' information disclosure. We find that many users reverse their decision when prompted with a fine-grained request.