Security Researcher Unmasks AI Chatbot Scam Inside Fake Dating App Dora

A security researcher analyzing a fraudulent dating app called Dora got a pop-up mid-investigation: an incoming call from a user named Jennifer. Her profile described a 41-year-old Sagittarius with red hair, blue eyes and piercings who liked music, horror movies, nightlife and sports. When researcher Matthew Gore-Kormanik, who goes by the handle Zigula, answered, the conversation that followed exposed how far romance scams have moved beyond scripted text.

The call did not come from a human. According to Gore-Kormanik's findings, Jennifer was an AI-generated persona built to keep targets engaged, mirror their interests and steer them toward payments. The app itself is fraudulent — there is no real dating pool behind it. Instead, the operation pairs chatbot technology with fake profiles to manufacture intimacy at scale, a combination that makes the old Nigerian-prince email look primitive.

Romance scams are already among the most costly categories of consumer fraud. The FBI has reported billions in annual losses tied to confidence fraud and romance schemes, with victims often drained through gift cards, crypto transfers and wire payments. What AI changes is the labor math. A single operator can now run hundreds of simultaneous conversations, each one tailored to the target's stated hobbies and emotional vulnerabilities, without hiring a call center or writing a single script.

The Dora case also shows how voice is becoming part of the con. Real-time synthetic speech lets a fake profile pick up the phone, which defeats the common advice to request a call before trusting an online match. Gore-Kormanik's account suggests the persona was convincing enough to sustain the interaction, and the app's design — pop-up calls, curated bios, instant messaging — is tuned to feel like a mainstream dating product rather than an obvious lure.

For users, the practical takeaway is that voice and video verification no longer prove much on their own. For platforms, the pressure is mounting to detect coordinated fake accounts and synthetic media before they reach victims, a task app stores have largely left to overstretched trust-and-safety teams. Regulators, meanwhile, have focused on the payment rails scammers use rather than the AI tools that start the relationship.

What remains unclear is how many Dora-style operations are running and who is behind them. Gore-Kormanik's work offers a rare look inside one, but the techniques are cheap and widely available, meaning the next version of this scam may not need a fake app at all — just a cloned face and a stolen conversation history.