This paper suggests a framework for initiating the construction of an adaptive global intelligent system for the detection of terrorist activities. The proposed system's main features are based on already known recommendations and emerging technologies. However, this research concentrates on the missing elements of collaboration. The paper presents the technologies and principles on which this framework will be based, while describing its relevant vision. The discussion is divided into the following categories: distributed information and data gathering while preserving privacy constraints; adaptive collaboration of software agents and domain-specific language; and secured network topology based on modular peer-to-peer and web services.