AI is Supercharging Romance Scams with Deepfakes and Bots
Online dating scams are nothing new. They’ve been around for years, and tend to ramp up as Valentine’s Day approaches.
However, what is new is the use of AI by bad actors running the scams, which is making them more difficult to spot, and the industrial scale of the operations behind them as organized crime groups out of Asia and elsewhere run massive scam compounds that escalate the threats.
The use of AI and the expansion of scam compounds come against the backdrop of an online dating trend that continues to grow, with online dating site eHarmony saying last year that about 30% of the adults in the United States – about 80 million people – have used online dating sites or apps.
The environment is ripe for online romance scams. According to the 2026 Norton Insights Report: Artificial Intimacy, almost half of online daters in the United States have been targeted in a dating scam, with 74% of those becoming victims. In addition, Gen Digital’s Gen Threat Report found that in the fourth quarter last year, more than 17 million dating scam attacks were blocked, a 19% year-over-year increase.
Cybersecurity firm McAfee found that one in seven U.S. adults – about 15% – said they’ve lost money to an online dating or romance scam.
And the money lost is significant. The Federal Trade Commission said that reported losses to romance scams in 2023 hit $1.14 billion. McAfee wrote about a 40-year-old woman who was a healthcare professional who lost $80,000 in a scam.
An Evolving Scam Environment
“Romance scams are no longer run by smooth-talking individuals sitting behind laptops,” Alex Quilici, CEO of YouMail, which protects users from spam calls, told Security Boulevard. “They’ve evolved into industrialized fraud operations that combine emotional engineering with sophisticated technology. The people running these scams know exactly how to build trust, how to mirror your communication style, and how to escalate the relationship until you’re all-in, emotionally and financially.”
Fueling the problem is technology.
“We’re seeing scammers use AI to write messages, draft backstories, and even clone voices,” he said. “When a victim hears what sounds like a real person, someone who claims to care about them, it’s incredibly hard to push back. That’s why we tell people: assume emotional manipulation is part of the playbook. If a new online relationship suddenly feels too perfect, too fast, or too intense, that’s a signal.”
AI Deepfakes are Changing the Game
Blackbird.AI’s Constellation Platform is an AI-driven narrative intelligence platform that helps organizations and governments detect and mitigate online misinformation, disinformation, and manipulative narratives. AI-powered deepfakes have “revolutionized the way scammers operate in the online dating world,” according to Emily Kohlman, senior intelligence analyst with the company.
“AI tools assist scammers in creating compelling online dating profiles,” Kohlman wrote in a blog post. “These profiles often feature deepfake-generated images and videos made with face-swapping technology, using pictures stolen from real individuals or created entirely by AI. Scammers also employ AI chatbots to engage in conversations, slowly building a connection with the target to manipulate them into financial transactions.”
The technology has made many of the tools people would use to verify that the photo of the person they were looking at was real, according to Roger Grimes, CISO advisor for KnowBe4.
Face-Swapping Tools and Automated Bots
“For years, the gold standard for verifying an online match was the custom photo request: ‘Send me a selfie holding today’s newspaper,’” Grimes wrote. “In 2026, that test is dead. Scammers can now instantly generate an image of themselves in any location or holding any object. Media alone is no longer proof of identity.”
The use of AI goes beyond static images, he added. Bad actors armed with AI-powered face-swapping and voice synthesizing tools can create deepfake video calls, while automated bots can maintain deeply emotional and visually convincing relationships for months, with the victim believing they are interacting with a human.
In addition, AI lets scammers pose as famous figures to further the connection with their victims.
From Initial Contact to Stolen Money
Romance scams unfold in fairly consistent ways. The scammer initially contacts the target online through a text supposedly sent due to a wrong number, a dating match, a reply to a comment, or some other ruse, and then escalates to daily messages that quickly turn intimate, according to McAfee.
From there, the scammer suggests moving the conversation to a private channel like WhatsApp or Telegram, then builds credibility by talking about a job they have, sending photos or AI-assisted videos. Eventually, the financial request comes – a small emergency, a cryptocurrency investment opportunity, gift cards, a plane ticket, or something similar – followed by a demand for something more, such as a verification code, co-signing something, or opening an account, which can lead to an account takeover or identity theft.
Once they have the money, the scammers disappear, ending communication and deleting accounts.
Don’t Send Money
“AI reduces the ‘tells’ that used to give scammers away,” McAfee wrote. “Deepfake audio and video can make someone appear real-time credible. Bot-driven chat can sound polished, attentive, and emotionally responsive.”
Among the clues that people reported seeing that told them they were dealing with a bot or fake profile were responses that felt scripted or repetitive, or were sent instantly or flawlessly, photos that appeared unnatural or AI-generated, and the person avoiding voice or video calls.
Security pros advise against sending money, crypto, gift cards, or anything else to someone you haven’t met in person or buying into promises of big investment returns. They also suggest doing reverse image searches of the person’s profile photos.
