AI technologies are applicable to Homeland Security, particularly in problems of analysis, discovery and prediction. We describe relevant AI techniques including Link Analysis, dynamic Decision Support and Agent Organization modeling, whereby anomalous activities may be detected and threats deterred. We also discuss application of techniques in local government and to security matters close to home. We conclude that AI Technologies are immensely useful for improving security of citizens and the homeland, expecially in conjuction with skilled human experts. But our position is that privacy cannot be fully protected if security is to be substantially enhanced.