Artificial Intelligence for Researching and Investigating Organized Crime

• In Event: WG Organized Crime and Criminal Networks Panel 1. Methods for studying organized crime

Abstract

The introduction of Artificial Intelligence in law enforcement has shown promise but concerns loom regarding its potential contributions to the proliferation and sophistication of crimes. This awareness brings the law enforcement community to a conclusion that the use of AI could be a benefit for criminals but also for law enforcement.
This paper aims to delineate the potential benefits and limitations of leveraging generative AI in researching and investigating organized crime (OC) and financial crime. By exploring both aspects, law enforcement agencies can optimize their investigative efforts to develop strategic and tactical intelligence.
The conceptual framework employed aligns with the one developed by Transcrime for the UNODC and INEGI Center of Excellence in 2012(https://www.transcrime.it/wp-content/uploads/2014/05/CoE_MOC-in-Latin-America-and-the-Caribbean-Sept-2012.pdf) This framework encompasses five dimensions (Groups, Activities, State Response, Enablers, Civil Society Response) measurable through quantitative and qualitative data at Macro (territory), Meso (vehicle), and Micro (subjects) levels. The main assumption is that OC risks vary across different socio-economic contexts, organizational structures, and vulnerabilities, forming a basis that generative AI can utilize to map risks related to OC and financial crimes. Examples of the methodology and its implementation are provided.

Author

• Ernesto Savona, Transcrime / Università Cattolica del Sacro Cuore (Milan)