A man in Shanghai, China, has been defrauded of nearly $28,000 after falling prey to an AI-enabled romance scam, according to reports from Chinese state media. The incident underscores the alarming trend of scammers leveraging generative artificial intelligence to create highly convincing fake personas. The victim, whose identity has not been disclosed, became romantically involved with a fictional woman named 'Ms. Jiao,' whose entire existence—including realistic videos, photos, and personal details—was fabricated using advanced AI tools.
The scammers used generative AI programs to produce deepfake videos and images of a young woman who never existed. After the victim developed romantic feelings for this virtual character, he transferred approximately 200,000 yuan (roughly $28,000) to a bank account controlled by the fraudsters. To make the deception more credible, the criminals created a fake identity for Ms. Jiao and even forged medical records and bills, claiming she needed urgent financial help. The emotional manipulation typical of romance scams was amplified by these AI-generated assets, making the fake persona seem real and desperate.
The Rise of AI in Romance Scams
Romance scams have been a persistent threat for decades, but the integration of artificial intelligence has dramatically escalated their sophistication. AI-powered tools allow scammers to generate lifelike faces, voices, and even video calls that can fool unsuspecting victims. Unlike older scams that relied on stolen photos or scripted messages, modern fraudsters can now create entirely new identities on demand. This technology lowers the barrier for scammers while raising the difficulty for victims to verify authenticity.
The use of deepfakes—AI-generated synthetic media that replaces a person's likeness—is particularly dangerous. In the Shanghai case, the scammers likely used a Generative Adversarial Network (GAN) to produce high-quality videos of Ms. Jiao speaking, smiling, or performing everyday actions. These videos can be customized to mimic real-time interactions, such as nodding or lip-syncing to scripted audio. Additionally, AI can generate text messages and voice notes that sound natural, further blurring the line between real and fake.
Widespread Impact of AI-Driven Fraud
The problem extends far beyond a single incident. A report released by cybersecurity firm McAfee on February 11, 2025, reveals that 52% of people have been scammed out of money or pressured to send money or gifts by someone they met online. The company notes an 'explosion of online romance fraud' through social media platforms, messaging apps, and AI chatbots. According to the study, 26% of individuals say they—or someone they know—have been approached by an AI chatbot posing as a real person on a dating app or social media. Furthermore, 21% have been contacted by someone pretending to be a well-known public figure. Of those who fell for it, 33% lost money, with an average reported loss of $1,985.
McAfee also blocked 321,509 fraudulent URLs designed to lure victims in the seven weeks leading up to Valentine’s Day 2025. These websites often mimic legitimate dating platforms or financial aid services, capturing personal information and payment details. The scale of this activity indicates a well-organized cybercriminal ecosystem that continuously adapts to new technologies.
Another high-profile case emerged in early 2025 when a woman in France was duped out of €830,000 ($850,000) in an AI scam where she believed she was in a relationship with Hollywood actor Brad Pitt. Scammers used deepfake videos and AI-generated voice calls to impersonate the star, creating an elaborate ruse that lasted months. The victim transferred large sums under the pretense of medical expenses and legal fees for the fictional relationship.
How AI Fuels the Fraud Ecosystem
Generative AI tools, such as text-to-image models like DALL-E and Midjourney, as well as voice cloning software, have made it easier for scammers to produce convincing content at low cost. In the past, creating a fake identity required manual photo editing or hiring actors, which was both expensive and risky. Today, a single prompt can generate thousands of unique images, each with consistent facial features. Voice cloning can replicate a specific person’s speech patterns using just a few seconds of audio, often harvested from social media videos.
These technologies are often combined with social engineering tactics. Scammers build trust over weeks or months by exchanging messages, phone calls, and eventually video chats—where AI-generated avatars or pre-recorded deepfake clips are used. Victims may be asked to send money for travel, medical emergencies, or business opportunities. Once the money is transferred, the scammers disappear, often using cryptocurrency or anonymous payment methods to avoid traceability.
Psychological Manipulation and Red Flags
The emotional toll on victims is severe. Romance scams exploit loneliness, empathy, and the desire for connection. In the Shanghai case, the creation of fake medical records played directly on the victim’s compassion. Scammers often isolate their targets from friends and family, making them more dependent on the fake relationship. They may also claim to be overseas, in danger, or facing legal trouble to solicit funds quickly.
Experts recommend several red flags to watch for: an online love interest who refuses to meet in person or video chat in real time; requests for money for emergencies, travel, or investments; and inconsistent stories that change over time. With AI, even video calls can be faked, so additional verification steps are essential. Reverse image searches can sometimes identify stolen photos, but deepfakes may not appear in any database. Using dedicated background check services or insisting on a live video call with specific gestures can help, though scammers may still find ways to cheat.
Legal and Regulatory Responses
Governments and law enforcement agencies are increasingly aware of the threat. China has implemented strict regulations on AI-generated content, requiring deepfakes to be labeled as such. However, enforcement remains challenging, especially when scammers operate across borders. In the United States, the Federal Trade Commission (FTC) has stepped up efforts to combat romance scams, and the FBI’s Internet Crime Complaint Center (IC3) tracks related complaints. In 2023, the FTC reported that romance scams led to losses of over $1.3 billion in the U.S. alone.
Meanwhile, tech companies are developing detection tools. McAfee, for instance, uses AI to identify deepfake videos and fraudulent websites. Social media platforms are improving account verification and banning known scam profiles. However, the cat-and-mouse game between scammers and defenders continues, as new generative models emerge that can evade existing safeguards.
Global Statistics Highlight a Growing Crisis
Beyond the McAfee report, other studies paint a grim picture. A 2024 study by the Global Anti-Scam Alliance found that online scams cost victims worldwide over $50 billion annually, with romance scams accounting for a significant portion. The average loss per victim in romance scams exceeds $2,000, but high-profile cases like the Shanghai and French incidents show losses can reach hundreds of thousands.
Demographically, older adults are particularly vulnerable, but younger people are also targeted. The McAfee data indicates that 26% of individuals approached by AI chatbots on dating apps suggest that younger, tech-savvy users are not immune. Scammers adapt their tactics based on the platform, using LinkedIn for professional networking scams, dating apps for romance fraud, and social media for impersonation.
Protecting Yourself in an AI-Driven World
Awareness is the first line of defense. Consumers should be skeptical of online relationships that progress too quickly or involve financial requests. Using video calls with real-time interaction, rather than pre-recorded messages, can help. Checking for inconsistencies in stories and verifying a person’s online presence across multiple platforms are also useful strategies. For those concerned about deepfakes, tools like Intel’s FakeCatcher can analyze videos for signs of manipulation, though these are not widely available to the public.
Financial institutions are also stepping up. Many banks now flag unusual transfers to new accounts or high-risk destinations. Victims should report scams immediately to their bank and local authorities, as early reporting increases the chance of recovering funds. In China, the victim likely notified the police, but recovering money from international scammers is difficult.
The McAfee report recommends that users never send money or gifts to someone they have not met in person, and to be wary of any request for financial assistance. Additionally, enabling two-factor authentication on social media and dating accounts can prevent scammers from hacking genuine profiles.
As AI technology continues to advance, the sophistication of romance scams will only increase. The Shanghai case serves as a stark warning. While generative AI offers incredible benefits for creativity and productivity, its misuse by criminals demands constant vigilance and updated countermeasures. The emotional and financial damage inflicted on victims is profound, and society must work collectively—through education, legislation, and technology—to mitigate this growing threat.
Source: ReadWrite News