Crime-busting AI detects fake robbery claims

Folake Dosu
November 2, 2018
Updated: April 29, 2019
Folake Dosu
November 2, 2018
Updated: April 29, 2019

AI-crime-fake-robbery

Artificial intelligence is extending the long arm of the law. An AI system called VeriPol is aiding the interrogations of law enforcement agencies in Spain, ZDNet reports. The tool, jointly developed by Cardiff University and Charles III University of Madrid, is helping to uncover false statements and fake crime and theft claims. 

Such fraud is a bane to insurance providers, as these companies are often the unwitting benefactors of criminals who cite illegitimate crimes of robbery in order to collect settlements through their policies.

The program works by reviewing and analyzing written statements in order to spot patterns that trend highly in falsely given statements. Some of the characteristics of these dubious claims are: descriptions of the criminal perpetrators, details surrounding the incidents and even the type of items listed as stolen. ZD reports that fake reports usually lack supporting details of the event such as the clear descriptions of attackers, and tend to focus on the items allegedly stolen and not the circumstances of the incident itself. Phony claims also trend towards having a lack of witnesses and delays between the robbery and notification of law enforcement. 

VeriPol uses what is called natural language processing, an element of machine learning technologies which helps artificial systems understand and interpret the natural use of human language. Algorithms are used to decode this language, and VeriPol's understanding is based on historical police reports which have been fed into the system.

Which items are the most reported in false claims? Samsung smartphones, iPhones and other high end tech items. Conversely, bicycles and jewelry are among the leading items reported in legitimate claims. “VeriPol uses what is called natural language processing, an element of machine learning technologies which helps artificial systems understand and interpret the natural use of human language. Algorithms are used to decode this language, and VeriPol's understanding is based on historical police reports which have been fed into the system.”

Researchers from the universities believe VeriPol has a major role to play in alleviating the stress and strain investigating false robbery claims puts on resources and budgets of many law enforcement agencies. But according to researcher Dr. Camacho-Callados, it could prove to be an effective deterrence. Falsely reporting a crime is a crime in itself and clocking the program at over 80% accuracy, researcher Dr. Camacho-Callados is hoping VeriPol can prevent that crime before it is even committed.

Great Companies Need Great People. That's Where We Come In.

Recruit With Us