International Association
for Cryptologic Research

Third-party apps enable a personalized experience on social networking platforms; however, they give rise to privacy interdependence issues. Apps installed by a user's friends can collect and potentially misuse her own personal data inflicting \textit{collateral damage} on the user herself while leaving her without proper means of control. In this paper, we present a study on the \textit{collateral information collection} of apps in social networks. Based on real data, we compute the proportion of exposed user attributes including the case of profiling, when several apps are offered by the same provider.