A man in Shanghai, China, has been defrauded of nearly $28,000 (approximately 200,000 yuan) after falling victim to an intricately crafted AI romance scam, according to Chinese state media. The scammers used generative artificial intelligence to create hyper-realistic photos and videos of a young woman they called 'Ms. Jiao.' Once the man developed romantic feelings for this fictional persona, he was repeatedly manipulated into transferring money, ultimately losing his life savings.
This incident is not isolated. As artificial intelligence technologies become more accessible and sophisticated, fraudsters are increasingly weaponizing them to perpetrate romance scams with unprecedented realism. Unlike traditional catfishing—which relies on stolen photos and fabricated details—AI-generated scams allow criminals to manufacture entirely new identities complete with lifelike images, voice recordings, and even real-time video calls. The Shanghai victim, for instance, was shown fabricated medical records and hospital bills that depicted Ms. Jiao as a woman suffering from a serious illness, compelling him to provide financial assistance out of love and concern.
The Rise of AI-Enhanced Fraud
The case underscores a broader trend in online deception. On February 11, 2025, cybersecurity firm McAfee released a report revealing that more than half (52%) of individuals have been cheated out of money or pressured into sending gifts or cash by someone they met online. The company has observed an 'explosion of online romance fraud' across social media platforms, dating apps, and AI chatbots. According to the study, 26% of respondents said they—or someone they know—had been approached by an AI chatbot impersonating a real person on a dating site or social network.
Additionally, 21% of people reported being contacted by someone pretending to be a well-known public figure. Among those who fell for such impersonations, 33% lost money, with an average reported loss of $1,985. In the seven weeks leading up to Valentine's Day 2025, McAfee blocked 321,509 fraudulent URLs specifically designed to lure romance scam victims. These links often lead to fake profiles, bogus dating sites, or phishing pages that harvest personal and financial information.
Anatomy of a Modern Romance Scam
The Shanghai scam followed a now-familiar pattern: the scammers first established a fictional identity using generative AI tools like Midjourney or DALL-E to create attractive, realistic images. They then used AI video synthesizers to generate short video clips of 'Ms. Jiao,' making her appear alive and present during text conversations. To deepen the illusion, they created a complete backstory—her name, occupation, family background—and even forged medical documents using generative design software. The victim, believing he was in a genuine relationship, was gradually persuaded to send money for medical expenses, travel costs, and other fabricated emergencies.
Such manipulation exploits emotional vulnerabilities. Romance scams are particularly devastating because they combine financial ruin with psychological trauma. Victims often feel shame, guilt, and isolation, which discourages them from reporting the crime. The Federal Trade Commission (FTC) in the United States reported that romance scam losses exceeded $1.3 billion in 2023, and AI is poised to drive that number higher. In January 2025, a French woman lost €830,000 ($850,000) after being convinced she was in a relationship with a deepfake version of actor Brad Pitt. The criminals used AI-generated videos and voice calls to maintain the ruse for more than a year.
How AI Makes Scams More Believable
Generative AI has democratized the ability to create convincing synthetic media. Modern tools can produce images with perfect lighting, consistent facial features, and natural backgrounds. Voice cloning software requires only a few seconds of audio to replicate someone's speech patterns. Deepfake technology can superimpose faces onto real-time video feeds, enabling scammers to impersonate anyone during live calls. For romance scammers, this means they can adapt their tactics to match victims' expectations—sending unique photos, speaking with a specific accent, or making video calls that appear authentic.
Moreover, AI-powered chatbots can maintain long conversations without human fatigue. These bots are trained on vast datasets of romantic dialogues, allowing them to express empathy, share stories, and build emotional bonds over weeks or months. Some advanced bots can even detect a victim's mood and adjust their responses accordingly. The McAfee report noted that 26% of people had been approached by an AI chatbot on a dating app, and many did not realize they were interacting with a machine until after they had already shared sensitive information or sent money.
Historical Context and Escalating Threats
Romance scams are not new. The first documented cases date back to the 18th century, with lonely-hearts ads and postal correspondence. In the digital age, they migrated to chat rooms, then to social media and dating apps. However, the introduction of AI has transformed the scale and sophistication of these operations. Previously, scammers had to manually craft messages and manage multiple fake profiles—a time-consuming process that limited their reach. Today, a single fraud ring can deploy hundreds of AI-generated personas across different platforms, each capable of targeting multiple victims simultaneously.
Law enforcement agencies are struggling to keep pace. Chinese police have arrested several suspects linked to the Shanghai case, but the perpetrators often operate from jurisdictions with lax cybercrime laws. Cross-border investigations are slow, and AI-generated evidence is difficult to authenticate. Meanwhile, the scammers' methods continue to evolve. In 2024, researchers at Sophos discovered a new breed of romance scam that uses AI to generate entire fake social media histories, including posts, comments, and friend networks, making the persona appear fully integrated into a community.
Protecting Yourself from AI Romance Scams
While technology makes scams more convincing, certain red flags remain consistent. Be wary of individuals who refuse to meet in person or video call in real time, especially if they offer excuses about poor internet connections or camera problems. Reverse image search any photos to check if they appear elsewhere on the web. Be suspicious if a new online acquaintance quickly professes deep affection and begins asking for money, even if they supply compelling stories, photos, or documents. Never send funds to someone you have not met face-to-face, especially via untraceable methods like wire transfers or cryptocurrency.
Cybersecurity experts recommend using AI detection tools, such as those that analyze metadata or subtle artifacts in images and videos. However, these tools are not foolproof. The most effective defense is skepticism: if an online relationship progresses too fast or feels too perfect, it may be a fabrication. Education is also crucial. Public awareness campaigns like the FTC's 'Pass on the Pics' initiative help consumers recognize the hallmarks of romance scams. As AI capabilities grow, so must our collective vigilance.
The Shanghai victim's story is a cautionary tale for the digital age. The man, whose identity has been withheld by authorities, reported the scam after realizing the woman he loved never existed. Police are investigating, but recovering the lost funds is unlikely. In the meantime, AI-generated romance scams continue to proliferate, exploiting human loneliness and trust. The McAfee report's findings—that 52% of online daters have faced financial pressure—serve as a stark reminder that the romance industry's digital transformation carries hidden dangers. As one expert noted, 'The heart wants what it wants. But AI can mimic that desire with terrifying precision.' Without robust safeguards and personal caution, more victims will inevitably fall into the arms of a ghost.
Source: ReadWrite News