A woman in Chengdu stumbled upon a matchmaking agency video on a social platform, only to lose 2,000 yuan in service fees before the promised "male client" stopped answering her calls and the agency staff went silent. This single complaint triggered a full-scale police investigation that ultimately uncovered a sophisticated fraud ring using artificial intelligence to fabricate the perfect male partner.
This money-making scheme promised wealthy, successful husbands—men like a 67-year-old "Uncle Li" with an annual income of 6 million yuan and properties in both Hangzhou and Shanghai, or a 1965-born "Mr. Liu" with assets of around 100 million yuan and homes in Germany and Shanghai. The carefully crafted profiles depicted men who were generous, willing to pay bride prices, buy homes, and cover daily household expenses, with all the details designed to reel in women aged 45 and above who were eager for companionship.
The Chengdu High-tech Public Security Bureau has now successfully cracked this case, arresting 28 suspects who had defrauded more than 50 victims of over 1 million yuan in less than a year. The operation, known as "Sichuan Wanhe Xitang Matchmaking Service Company," was a fictional entity established in October 2025, running parallel scams in both Chengdu and Kunming, Yunnan Province. What made this operation especially insidious was its deceptive layer of sophistication: the company maintained different registered names at different locations—"Wanhe Xitang" at one address and "Wanjin Xitang" at another—creating confusion for investigators.
The scam worked through a systematic process. First, the group used live-streaming on social platforms with professional scripts targeting women aged 50-70 looking for spouses. They deployed professional editing teams to use AI technology to generate massive numbers of fake images depicting successful middle-aged and elderly men, complete with elaborately crafted biographical experiences. To further cement trust with victims, streamers would even use voice-changing devices to pose as these "high-quality men" and conduct live calls, creating a facade of genuine interest that convinced on-screen viewers of the authenticity of the profiles.
Once a victim expressed interest through private messages, online "matchmakers" would redirect them to WeChat, citing platform privacy concerns. After displaying the fabricated male profiles, they would claim that the "gentleman" was already impressed and that payment of the matching service fee would facilitate immediate contact. The victims were asked to choose a service package and pay upfront—the only way, they were told, to be introduced to their ideal match. But police discovered that the "male clients" were purely digital creations, with a single AI-generated face being recycled across multiple identities depending on which victim was being targeted. Behind the same photo, the man would be "Mr. Li" to one woman and "Mr. Wang" to another.
After receiving payment, the company would assign staff to send messages from fabricated social accounts, engaging victims in conversation until, at a predetermined moment, they would terminate contact citing "personality differences" or some other incompatibility. Meanwhile, each link of the criminal chain—from initial outreach through sales and after-service to the final editing and streaming—was compartmentalized and run by different teams located in different cities.
The mastermind, a man named Wang Mouxin, worked remotely from Kunming, where he oversaw the sales and after-service teams. His editing and streaming crews operated from Chengdu. This geographic separation was a deliberate strategy to avoid detection, making the operation exceptionally difficult to trace using conventional investigation methods.
On July 2, Chengdu's High-tech Public Security Bureau coordinated a cross-provincial sweep with Yunnan police, arresting all 28 suspects in a synchronized operation. Even as police moved in, the scam group was still live-streaming with their AI-generated male personas. The subsequent raid uncovered a trove of "closing tactics" scripts and meticulously designed scripts for conversation. From the development of AI personas to scripted interactions and live broadcasts, every element had been carefully choreographed to maximize profit.
The investigation revealed a concerning trend: the primary suspects are predominantly "post-90s" generation members, with some even from the "post-00s" generation—individuals who are familiar with AI technology but chose to pervert it for criminal purposes. All victims were single women aged 45 and above.
This case shows how modern fraud rings have evolved with technological advances. The group's success relied not just on deep-fake imagery but also on sophisticated sales scripts developed by young, tech-savvy criminals who operated the business like a startup, complete with defined departmental roles. The 28 suspects are currently in criminal detention, and the case remains under further investigation. Police continue to emphasize their commitment to cracking down on new types of crime and protecting public property safety.