Social networking sites (SNSs) are becoming increasingly popular on the Web. Sharing personal information in these networks can be dangerous, considering that malicious users can get access to this information and use them for purposes other than the original. Although SNSs typically provide tools for users to set who can access their shared data, this access restriction only applies to network users and not for third parties and the social network itself. In this paper, we present both a mechanism to enhance privacy in SNSs and taxonomy for classifying SNSs privacy policies. We combine and extend two taxonomies of privacy, unifying them to classify SNSs privacy policies and also the aforemen-tioned mechanism. We evaluated the taxonomy classifying six SNSs privacy policies and the mechanism, presenting the results and our findings about the classification.