Cross-Border KYC AI Identity Document Forgery Scam: Using AI-Generated Face Videos to Fraudulently Obtain Financial Access

Who Gets Targeted

Victims are mostly cross-border e-commerce sellers, freelancers, and small and medium-sized business owners. They urgently need to quickly open cross-border payment collection accounts but lack professional compliance teams, making them easy targets for low-price ads such as “AI-accelerated cross-border KYC” and “approval in minutes.” They place blind trust in AI technology, seek convenience and low fees, readily believe anonymous links sent via social software, and hand over the front and back of ID cards and selfie videos to platforms of unknown origin. In terms of losses, payment accounts opened using stolen identities are used to transfer fraud- and gambling-related funds; funds transferred in a single case reach more than RMB 100,000 (media estimate, not independently verified). In similar cases where AI-forged faces bypass liveness authentication, victims’ bank cards were fraudulently debited more than RMB 50,000 (media estimate, not independently verified). Victims often do not notice until their accounts are frozen and their identities continue to be misused.

骗局怎么运作

• Step 1: Scam groups post “AI-accelerated cross-border KYC” ads on social platforms such as Telegram and WhatsApp, claiming that through self-developed deep face-swapping technology they can complete identity verification in minutes, and attract business owners or individuals urgently needing to open accounts with fees as low as several hundred yuan.
• Step 2: The victim clicks a link and enters a login page disguised as a legitimate KYC platform. The platform asks them to upload the front and back of their ID card and a selfie video, then tells them that “AI-enhanced face matching” is required. Without knowing it, the victim uploads real materials, and the system backend uses an AI model to replace the victim’s facial information with a pre-prepared, highly similar synthetic video.
• Step 3: The forged video, together with the uploaded ID information, is submitted to the target cross-border financial institution’s automated KYC system. Because the system mainly relies on machine learning-based facial similarity assessment and lacks liveness detection and multimodal comparison, the synthetic video can easily pass review, and the system returns “identity verification passed.”
• Step 4: The scammer uses the verified account to immediately activate cross-border remittance, cryptocurrency trading, or virtual asset custody services, completing large-value transfers or purchasing virtual goods in the victim’s name, with cumulative cash-out amounts ranging from several hundred thousand yuan to tens of millions of yuan.
• Step 5: Several days later, the victim receives a risk alert or account freeze notice from the financial institution, discovers that the account has been flagged as abnormal, all funds have been transferred to overseas wallets, and the real identity information has been stolen for subsequent larger-scale identity fraud.

红旗信号（看到这些快跑）

• 🚩 Asks for video rather than a static photo for identity verification, and claims that using AI can improve the pass rate
• 🚩 Fees are significantly lower than the standard for legitimate industry KYC services, and induce quick payment with “limited-time offer” and “instant approval”
• 🚩 Uses unofficial domains or sends links directly via social software, without providing legitimate registration or filing information
• 🚩 Promotional content contains technical terms such as “AI face swap” and “deepfake,” yet lacks any explanation of liveness detection
• 🚩 During verification, asks victims to download and install apps or plug-ins from unknown sources on their own in order to obtain camera permissions

真实案例

• In December 2025, a large-scale account theft incident occurred in a user group of the AI chat platform Claude. Hackers used AI face-swapping technology to forge users’ faces and completed fund transfers through the cross-border payment KYC process, involving about RMB 5 million (source: https://www.t00ls.com/articles-75498.html)
• In June 2026, a cross-border financial services company reported that its KYC system was deceived by AI-forged videos. Attackers used synthetic videos to complete 10 cross-border remittances totaling about USD 3 million (source: https://stock.10jqka.com.cn/20260817/c679017619.shtml)
• In March 2023, Fuzhou’s first case of illegal account authentication using AI face-swapping: two suspects used AI-generated face videos to intrude into a bank system, illegally opened accounts and cashed out about RMB 500,000, and were sentenced to fixed-term imprisonment by the court (source: https://news.fznews.com.cn/fzyw/20260317/69b8a05c53e24.shtml)

Official Stance

• Cyberspace Administration of China, Ministry of Public Security, 2025-03-21, issued the “Measures for the Security Management of Face Recognition Technology Applications” (source: https://www.cac.gov.cn/2025-03/21/c_1744174262156096.htm)
• CCTV.com, 2025-09-20, published “Cyber Police Crack Case of Illegal Intrusion into Computer Information Systems through ‘AI Face Swapping’” (source: https://news.cctv.com/2025/09/20/ARTICBh1QeKuSKTcLh7YxLEZ250920.shtml)

How to Protect Yourself

• ✅ Use multi-factor authentication; in addition to face recognition, a mobile phone verification code or hardware token must be bound
• ✅ Choose KYC service providers with liveness detection, ensuring that liveness actions such as blinking and turning the head appear in the video
• ✅ Compare uploaded ID card information with live real-person photos taken on site, avoiding reliance solely on algorithmic matching
• ✅ Regularly audit the AI models and face comparison algorithms used to ensure they have the ability to detect deepfakes
• ✅ Download and use KYC platform apps or webpages through official channels, and strictly guard against verification through anonymous links sent via social software