China: AI deepfake fraud and impersonation schemes documented
Investigative reporting in China documents widespread misuse of AI deepfake technology for fraud, impersonation and commercial exploitation. Cases include synthetic videos impersonating a doctor to sell products, technology being sold online with minimal controls, and verified fraud schemes using video deepfakes to trick people into transferring money.
What happened
A famous Chinese doctor Zhang Wenhong discovered an AI-synthesized video impersonating him promoting a protein bar product. The video prompted online sales of approximately 1,200 units before being removed.
A video demonstrated AI deepfake technology creating a synthetic video of a girl asking her parents for 20,000 yuan from a hospital, created from footage of her having a normal conversation with a friend.
AI face-swapping software is being sold online through multiple platforms with pricing from a few yuan to over 100 yuan. Sellers offer memberships (199 yuan for 2 years) and reseller opportunities (499 yuan commissions) to generate AI clone videos.
A Tianjin man was defrauded of 950,000 yuan across three transfers after receiving a video call from someone using AI-synthesized video appearing as his work supervisor. Police recovered 830,000 yuan.
Reported impact
The synthetic video of doctor Zhang Wenhong resulted in mass product sales based on fraudulent endorsement. Zhang stated the problem persists: 'These fake AI synthetic messages are like a locust plague, happening again and again, swarming like locusts flying across the sky to harm people.'
AI deepfake technology is being exploited for fraud targeting middle-aged and elderly people, with reported cases including scams impersonating billionaire Jack Ma on social platforms and actor Jing Dong on short video platforms using resource constraints as pretexts.
Index analysis
Expert analysis notes that unauthorized use of AI face-swapping to impersonate others may constitute violations of portrait rights as civil law violations and could constitute criminal liability including false advertising and online fraud depending on severity.
Selling AI synthesis software without controls could constitute shared criminal liability if the software is intentionally provided to support criminal activity.
Governance controls
The video promoting Zhang Wenhong's product endorsement was removed from the posting account, and that account was permanently banned by the platform.
Experts recommend legislation clarifying the application scope and legal boundaries of deep synthesis technology, defining responsibilities of platforms, technology providers, content creators and users, and establishing multi-stakeholder coordination to combat AI-related crimes and underground industry chains.
Classification
Sources and evidence
This record was researched and written by the Index, based on the following original reporting.