Man convicted for using AI face-swapping software to bypass facial recognition and defraud 23 people in China
A man in China was convicted of purchasing 1.95 million personal information records and using AI face-swapping software to generate deepfake videos that bypassed facial recognition systems to access 23 people's financial accounts and fraudulently purchase mobile phones worth 15,996 yuan. The court sentenced him to 4 years and 6 months in prison and ordered damages and a public apology; one financial platform's security vulnerabilities were exposed and subsequently remedied.
What happened
Fu Mouli purchased over 1.95 million personal information records of citizens through online means. He then used this information with AI face-swapping software to generate deepfake videos and gain access to the facial recognition systems of at least one financial payment platform.
Fu Mouli accessed 23 victims' financial platform payment accounts using the AI-generated deepfake videos to authenticate as them via facial recognition. He altered payment passwords and bound mobile phone numbers for 5 victims, and fraudulently charged 2 mobile phones worth 15,996 yuan using the account of 1 victim. He attempted to breach facial recognition systems on multiple other financial platforms but was unsuccessful due to stricter security controls.
Reported impact
23 people had their financial accounts accessed and at least two mobile phone purchases totalling 15,996 yuan were made fraudulently using stolen payment accounts.
Index analysis
The procuratorate noted that one financial payment platform had security vulnerabilities and insufficient security measures, relying on facial recognition alone without additional layers of identity authentication. The platform has since completed remedial measures. A procuratorate official stated that adding an additional layer of identity verification would have prevented the fraudulent account access.
Governance controls
Following discovery of the security gaps, the procuratorate issued a legal risk warning notice to the financial payment platform. The platform subsequently implemented security improvements. Experts noted that leading-edge technologies in the industry can detect AI-generated deepfakes with over 95% accuracy, and some smartphone manufacturers have already implemented anti-forgery verification procedures for facial recognition systems.
Classification
Sources and evidence
This record was researched and written by the Index, based on the following original reporting.