People rely on a diverse personal network of friends and contacts to get trusted information; help filter and interpret information; and get referrals to other people. In practice, connecting people of mutual interests helps them to work together and share social resources to achieve common goals. While online social networking tools can help individuals be more productive, there are more problems concerning security and privacy. For example, highly private information can be more easily leaked and spread in social media. In addition, adverse outcomes may also occur at any time and spread much faster when malicious users spread rumors and misinformation. In this talk, I present some of our work along the two dimensions. First, I discuss the privacy issues in social analytics and how we address them in both Internet and enterprise settings. Specifically, I introduce SmallBlue, an info-social sensing, analysis and visualization system designed to unlock valuable collective intelligence within organizations. It has been successfully fostering collaborations of IBMers in over 70 countries. SmallBlue also enables us to advance our understanding of organizational social networks, such as the information spreading, social correlation and culture factors. Next, I present our social-analytics-based approaches to enhancing the security of social systems. The ability to influence is now democratized by social media platform. Yet, such influence is susceptible to the ever growing hacking of social interaction such as bots. We aim to understand and detect such social media behavior towards the goal of prevent adverse outcome. In particular, we have investigated the detection of anomalous information spreading in social media. Such anomalous information spreading could potentially be rumors or real emergent events, both of which are important to notify human analysts for further investigation.