The primary goal of a terrorist-organized data leak is to share private data with the rest of the world. Once acquired, U.S. Government personnel information can provide actionable information to use against the U.S. in the form of direct strike, blackmail, fraud, or impersonation. For special operators at numerous government organizations, this leakage or discovery of personal information can have devastating national and international consequences. Without a reliable way to rapidly identify, assess the severity of, and respond to potential leaks, the safety of special operations forces is at risk. In the Cloud-based SOCOM Scalable Man-Machine Identity Learning Environment (C-SMILE) project, we are presenting a next generation identity management system that will provide an integrated suite of scalable, high performance technologies and automated analysis tools. Relying on a strong foundation in probabilistic modeling and natural language processing algorithms, C-SMILE technology enables early detection of data leaks posted online and produces a quantified assessment of leak severity. Our goal in Phase I of this effort is to provide an assessment on the feasibility of applying these technologies and tools to the domain of identity risk management.