Trick Me If You Can: How AI Solved a Murder Case Using Biometrics

In April, employees of Money Station, a Chinese microfinance organization (MFO), got an alert from their AI-powered facial recognition system that turned out to point to a murder.

The system sent a notification to the employees' messenger: "The client in the photo has their eyes closed. Confirmation of the loan is required."

Attached to the message was a photo from the online verification process. The image showed a young woman with a face marked by severe bruising, and a bright scarlet stripe was visible on her neck. The AI system also noted a discrepancy: the loan application was submitted under a woman’s name, but the voice used during the verification was distinctly male.

The operators contacted the police and attached the biometric check data.

A Crime Uncovered

Based on the information from the system, the Chinese prosecutor's office approved the detention of the suspect. Investigators tracked the geolocation recorded by the mobile app and arrested a 29-year-old man named Zhang at his parents' farm in the countryside, where he was trying to burn the body of his ex-girlfriend.

On April 11, 2019, Zhang and his girlfriend argued over money. When she tried to end the relationship, he beat her and strangled her with a rope. He then rented a car, hid her body in the trunk and drove to the countryside. From there he posed as the dead woman: he messaged her employer to report that she was on "sick leave" and tried to take out a loan through Money Station’s online platform.

How AI Foiled the Killer’s Plan

Zhang was unaware of the advanced facial recognition technology embedded in Money Station’s mobile app. The system included several layers of verification, such as liveness detection, which requires users to blink to confirm they are alive and not a mask, photo, or video. When Zhang failed this step, the system flagged the anomaly and escalated the case to the Head of Security for further review.

Zhang had already withdrawn 30,000 yuan from his ex-girlfriend’s bank account through a mobile banking app. The loan application was what gave him away. It led to a police raid and, in the end, a long prison sentence.

The Role of Biometric Technology

Biometric.Vision’s liveness detection technology, used in Money Station’s system, can determine within seconds whether a live person is in front of the camera. If suspicious behavior is detected—such as a lack of eye movement—the system immediately alerts company operators for further investigation. This robust security measure not only helps prevent fraud but, as in this case, can aid in solving serious crimes.

A Safer Future Through Technology

The case shows what biometric checks can do beyond fraud prevention. A liveness check built into a lending app protected the platform and gave investigators the evidence that solved a crime.

Follow our updates for more stories about technology at work. To learn more about biometric technology, visit our website at Biometric.Vision.