A man in Shanghai, China, has been defrauded of nearly $28,000 (200,000 yuan) after falling victim to an AI-powered romance scam, according to Chinese state media reports. The scammers used generative artificial intelligence to create realistic videos and photographs of a young woman they called 'Ms. Jiao.' The victim, who believed he was in a genuine romantic relationship, transferred the money to what he thought was her bank account after being manipulated with fabricated medical bills and other personal documents.
This case is part of a broader surge in AI-enhanced romance fraud. As artificial intelligence becomes more accessible and sophisticated, scammers are leveraging it to create deepfake personas, generate voice calls, and even produce video messages that appear authentic. The technology allows fraudsters to maintain long-term deceptions with minimal effort, often targeting vulnerable individuals seeking companionship online.
The Anatomy of an AI Romance Scam
Romance scams have existed for decades, but AI has dramatically altered the landscape. In the Shanghai case, the scammers went beyond static photos: they used AI to generate dynamic videos of 'Ms. Jiao' smiling, waving, or appearing in everyday settings. Such content can be compiled from a few real images or even created entirely from scratch using generative adversarial networks (GANs). The realistic output makes it nearly impossible for victims to distinguish fantasy from reality.
Additionally, the fraudsters created a fake identity—complete with a fabricated social media presence, fake job history, and counterfeit medical records showing expensive treatments. These elements were used to prompt the victim to send money for emergencies. The psychological manipulation common to romance scams—building trust, creating a sense of urgency, and exploiting empathy—becomes far more potent when paired with visual and audio proof that seems undeniable.
According to cybersecurity firm McAfee, which released a report on February 11, 2025, more than half (52%) of people surveyed said they or someone they know have been scammed out of money or pressured to send money or gifts by an online romantic interest. The report highlights an 'explosion of online romance fraud' across social media, dating apps, messaging platforms, and even AI chatbots. 26% of respondents reported being approached by an AI chatbot posing as a real person on a dating site or social network, while 21% had been contacted by someone pretending to be a well-known public figure. Among those who fell for the impersonation, 33% lost money, with an average loss of $1,985. McAfee's systems blocked 321,509 fraudulent URLs in the seven weeks leading up to Valentine's Day 2025, each designed to lure victims into romance scams.
AI Enables Scale and Personalization
One key advantage AI provides to scammers is the ability to scale their operations. Traditional romance scammers had to invest time crafting individual messages and maintaining multiple personas. Now, large language models (LLMs) can generate convincing, personalized conversations at scale, adapting to each victim's responses. Chatbots can be programmed to express affection, share fabricated personal stories, and even argue or reconcile, all while mimicking human emotional nuance. This allows a single fraud ring to target hundreds or thousands of victims simultaneously.
Deepfake technology further amplifies the threat. In January 2025, a French woman was duped out of €830,000 ($850,000) after believing she was in a relationship with Hollywood actor Brad Pitt. The scammers used AI-generated videos and images of Pitt, along with voice synthesis, to convince her of the actor's romantic interest. Such cases demonstrate that AI can exploit even well-known public figures' images, making the scams harder to detect.
The psychological underpinnings of these scams are well understood by fraudsters. They often target older individuals, people who are lonely, or those who have recently suffered a loss. The AI persona offers companionship, validation, and a seemingly intimate connection. Once trust is established, the scammer introduces a crisis: a medical emergency, a legal problem, or a travel mishap that requires immediate financial help. The victim, believing they are helping someone they love, sends money—often multiple times before realizing the deception.
McAfee Data and Industry Trends
The McAfee report also found that many victims are reluctant to report the crime due to embarrassment or fear of stigma. This underreporting makes it difficult for law enforcement to track the true scope of AI romance scams. However, the company's data on blocked URLs and incident reports suggests the problem is growing exponentially. In 2024, global losses from romance scams were estimated in the billions of dollars, with AI-related fraud accounting for an increasing share.
Romance scammers are also adopting hybrid tactics. Some combine AI-generated videos with real-time voice cloning during phone calls, making the impersonation even more convincing. Others use AI to analyze victims' social media profiles, tailoring their approach to exploit specific interests or vulnerabilities. For example, a scammer might learn that a target loves classic movies and then fabricate a story about working in film production, complete with AI-generated behind-the-scenes photos.
Law enforcement agencies worldwide are struggling to keep pace. Traditional methods of tracing money flows or identifying fake profiles are less effective when the profile images are entirely synthetic and not linked to any real person. Moreover, many AI-scamming operations are based in countries with weak cybercrime laws or poor international cooperation. Victims often lose their life savings with little chance of recovery.
Prevention and Awareness
Experts recommend several steps to avoid AI romance scams. First, perform reverse image searches on any profile photos: while deepfakes can fool many systems, some AI-generated images still show subtle artifacts in the eyes, hair, or background. Second, be wary of anyone who quickly declares strong feelings and refuses to meet in person or video chat live. If they do agree to a video call, look for unnatural blinking, lip-sync delays, or audio glitches. Third, never send money to someone you have not met face-to-face, especially under urgent circumstances. If a supposed romantic interest asks for funds for a medical bill, travel expense, or visa fee, it is almost certainly a scam.
Social media platforms and dating apps are also beginning to implement AI detection tools. Some use machine learning to identify synthetic profiles based on behavioral patterns. Others require users to submit a short video selfie to verify identity. However, these measures are not foolproof, and scammers continuously adapt.
In the Shanghai case, local police have launched an investigation but have not announced any arrests. The victim, whose identity was not disclosed, is reportedly devastated both financially and emotionally. He now faces the difficult process of rebuilding trust and recovering from the psychological manipulation. The incident serves as a stark reminder that as AI becomes more integrated into daily life, the line between real and fake will only blur further. Awareness and caution are the best defenses against this evolving threat.
The McAfee report emphasizes that education is critical. Many people do not realize how convincingly AI can impersonate a human being, nor how easily personal information can be weaponized against them. Organizations are urged to include romance scam awareness in broader digital literacy programs, especially for older adults who may be less familiar with AI capabilities. Financial institutions are also being asked to flag suspicious transactions that match common romance scam patterns—such as repeated small transfers to accounts in different countries.
As AI-generated content becomes indistinguishable from reality, the onus falls on both technology companies and users to remain vigilant. The romance scam that began with a simple fake profile has evolved into a sophisticated criminal enterprise that leverages the very tools designed to connect people. The story of the Shanghai man is not isolated—it is a warning of a future where AI can manufacture love, trust, and betrayal with equal ease.
Source: ReadWrite News