An AI tool developed to uncover fake reports

Summary:

An AI tool is helping the police to uncover false statements, fake crime and theft claims with over 80 percent accuracy. 

Client: 

The Spain’s National Police Corps

Problem Statement: 

People file false robbery reports for various reasons, such as avoiding embarrassment or making insurance claims. Previously, the primary method for identifying these fake reports involved having experienced police officers review suspicious cases, but this approach was not always effective.

Results: 

• VeriPol outperforms human experts by 15 to 20 percent.
• A significant increase in the detection of false reports and saving time for the police.
• The tool identifies false robbery reports with over 80 percent accuracy.

AI Solution Overview:

The Spanish national police introduced a tool called VeriPol in police stations to help detect false complaints, such as individuals falsely reporting robberies. VeriPol uses natural language processing, a ML technology that helps artificial systems understand and interpret human language. Algorithms decode this language based on historical police reports fed into the system.

VeriPol’s success lies in its ability to identify specific patterns in the language of fake reports, such as repetitive modus operandi or a lack of detailed information about the incident.

References: 

1. VeriPol: AI Crackdown on Fake Robbery Reports. https://medium.com/@AIWorldBlog/veripol-ai-crackdown-on-fake-robbery-reports-71d025a42e97
2.  Spanish police plan to extend use of its lie-detector while efficacy is unclear. https://algorithmwatch.org/en/spain-police-veripol/
3. VeriPol computer tool studies text to identify false police statements. https://www.theengineer.co.uk/content/news/veripol-computer-tool-studies-text-to-identify-false-police-statements
4. Crime-busting AI detects fake robbery claims. https://builtin.com/artificial-intelligence/crime-busting-ai-detects-fake-robbery-claims

Industry: Public Services

Vendor: Cardiff University and Charles III University of Madrid

Client: The Spain’s National Police Corps

Publication Date: 2018