AI deepfakes of economists used in stock-trading fraud scheme in China
Fraudsters in China used AI deepfakes of economists 任泽平 and 叶荣添 in fake stock-trading videos on social media to lure investors into scams. The scheme directed victims through multiple links to fake broker apps where they were pressured to transfer funds to unregistered third-party accounts, violating Chinese securities law. Regulators reported handling hundreds of complaints and referring dozens of cases for criminal investigation.
What happened
Fraudsters used AI deepfake technology to create synthetic videos impersonating economists including 任泽平 (Ren Zeping) and stock analyst 叶荣添 (Ye Rongchan), claiming to offer high returns through stock trading services. These fake videos were distributed on social media platforms at night, directing viewers through multiple QR codes and links to download apps named 立方私域 and 企小信, which housed fraud schemes.
Once investors engaged, they were directed to join groups claiming to represent securities firms 东方证券 and 开源证券. Fraudsters posing as brokers promised returns of 50-60% monthly and directed investors to transfer funds to third-party accounts including 清远市沄集生物科技有限公司 and 石家庄旭发装饰材料有限公司, violating Chinese securities law which requires client funds be held in bank accounts registered to the brokerage.
The scheme involved multiple deceptive layers: fake credentials shown by self-styled "assistants" and "customer managers," repeated changing of QR codes and links to evade detection, and claims that secrecy was necessary to prevent interference by other traders. The group chats appeared to contain 2000+ members but actual membership and identities could not be verified.
Reported impact
广东证监局 reported handling 238 complaints of illegal securities activities in its jurisdiction over one year, with 33 cases referred as criminal leads and 4.55 million yuan in fines and confiscations imposed.
Multiple victims shared their experiences on social media. The fraud scheme targeted investors seeking quick returns, with the fraudsters exploiting psychological tactics including promises of daily rewards and gift redemptions.
Index analysis
The deepfake videos were created using public videos of the economists as training material, with AI voice cloning services available commercially that could replicate voice features at low cost from audio samples. The high visual fidelity of the deepfakes made them difficult for investors to distinguish from authentic content.
Governance controls
The schemes deliberately circumvented platform oversight by frequently updating QR codes every few days and requiring multiple link redirects before users could join new WeChat groups. The perpetrators operated entirely within closed app environments where group member counts and identities could not be verified.
When reporters verified the identity of the fraudster "customer manager 陈哲" through 中国证券业协会, the registered credential photo did not match the person in the video. 开源证券's official customer service acknowledged possible identity theft and confirmed the fake app "开源资管版" did not exist on their official website.
Classification
Sources and evidence
This record was researched and written by the Index, based on the following original reporting.