Police bust deepfake video fraud ring in China, 50,000+ synthetic faces recovered
Police in Qingdao, China arrested 14 suspects running a fraud operation that created over 50,000 deepfake videos by synthesizing AI with stolen personal information, then used the synthetic videos to bypass facial recognition during account creation on social platforms. The fraudulent accounts were sold to criminals for further schemes including wire fraud and illegal content posting.
What happened
Qingdao Jiaozhou police discovered a suspect selling verified social media accounts and personal information including names, ID numbers, and photos online.
A criminal ring led by Zhou operated a workshop that collected and resold personal information of citizens. Members purchased static ID photos obtained through hacking or black market purchases, then used AI technology to synthesize them into dynamic deepfake videos showing head movements and eye blinking.
The synthetic videos were designed to bypass facial recognition systems used in real-name account verification on social platforms. Once verified, the accounts—which had payment and streaming functions—were sold for 150 to 300 yuan each.
Reported impact
Police recovered over 500,000 personal information records. Accounts created using the synthetic videos were used by fraudsters for wire fraud, bounty scams, and posting illegal gambling and pornographic content.
The Zhou ring made approximately 600,000 yuan in profit from selling illegal accounts. Information brokers including Ren and Li sold hacker-obtained personal details for 2 yuan per record.
Index analysis
The ring operated a tiered supply chain: hackers obtained data, brokers packaged and resold it, and AI was used to create synthetic verification videos. Criminals used virtual currency transfers to evade law enforcement.
Governance controls
Police conducted multi-province arrests, seizing over 80 mobile phones and computers, and 60 cloud servers. Officers stated they would work with platforms to improve security defenses and penalize platforms that failed to meet legal protection obligations.
Classification
Sources and evidence
This record was researched and written by the Index, based on the following original reporting.