Finding the Missing Links: Using AI to Map and Disrupt Transnational Crime Networks

Law enforcement intelligence on criminal networks is often incomplete, with only partial segments observable. This session presents a framework for reconstructing spatiotemporal criminal networks by integrating intelligence with environmental and geographic data derived from remote sensing. It asks: How do we connect an event in one country to a coordinating event in another? The presenter will share a real-world application from a complex operational environment known for organized crime. Using network analysis, the model mapped interactions and identified previously unobserved logistical nodes. Rigorous validation by the National Police of Colombia confirmed that these predictions represent strategic nodes governed by three spatial strategies: avoidance, strategic attraction, and calculated trade-offs.

• Upon completion, attendees will be able to convert spatio-temporal intelligence data (such as radio intercepts) to build a chronological network that models and maps real-world criminal interactions in distant locations.
• Upon completion, attendees will be able to identify how a machine learning framework can be applied to this network data to predict and locate a criminal organization's previously unobserved logistical nodes and hidden connections.
• Upon completion, attendees will be able to validate the critical process of using operational intelligence to validate predictive models and how to use these newly validated insights to plan effective strategic interventions.