That Hot Stranger Sliding Into Your DMs? Probably 40,000 Lines of Code.

Full Episode Transcript

Some of these accounts have tens of thousands of followers. Thirty-one thousand. Eighty-eight thousand. And the person in the videos? He doesn't exist. Not a single frame of him is real.

Trusted by Investigators Worldwide

Run Forensic-Grade Comparisons in Seconds

Court-ready facial comparison reports. Results in seconds.

Get Started

7-day refund guarantee**

🎆 July 4th Sale: 50% OFF your first month — use code JULY426 at checkout · ends July 11

If you've ever gotten a flirty message from a

If you've ever gotten a flirty message from a stranger who looked a little too good — this story is about you. Because the newest romance scams don't start with "send money." They start with attraction. They start by making you want them to be real. According to reporting from Let's Data Science, synthetic accounts are posting short, silent dancing videos — no talking, no sound. And they're pulling in followers by the thousands. The question I kept coming back to is this — why silent? Why would a fake profile stay quiet?

The answer tells you everything. For years, the best way to catch a deepfake was the mouth. Detection tools listened for whether the voice matched the lips. When they didn't line up, you had your fake. So the people building these accounts did something clever. They stopped using audio entirely. No voice, no lip movement, nothing to check. They didn't stumble into imperfect fakes — they engineered around the detector on purpose. That's the shift. This isn't one bad video. It's a system designed to slip past the exact tools meant to stop it.

And it's working at a scale that's hard to picture. The Federal Trade Commission says Americans lost more than three billion dollars to romance scams last year. The year before, that number was closer to one point two billion. So it more than doubled — in a single year. That jump lines up almost perfectly with the moment these A.I. tools became free and easy for anyone to use.

Now, you might think — surely I'd spot a fake photo. Here's the part that stops people cold. Researchers tested this. They showed people A.I.-generated faces and real ones, and told them upfront they were being tested. People still couldn't tell the difference. Their accuracy fell below a coin flip. Worse than guessing. So the old advice — do a reverse image search — mostly doesn't work anymore. You can't find the "original" photo online, because there is no original. The face was generated from scratch.

The Bottom Line

There's a term for the long game here — "pig butchering." Scammers spend weeks building real emotional trust before they ever ask for a dollar. That used to take human effort. One scammer, a handful of victims at a time. Now A.I. runs those conversations start to finish. One operator can build trust with thousands of people at once. For fraud investigators, that breaks the whole playbook. The face is fake, the voice is fake, the photos lead nowhere. The only thing left to read is behavior — how fast they push, whether they'll ever meet in person, whether the story stays consistent across accounts. For the rest of us, it means the warmest message in your inbox might be the emptiest.

Here's what surprised me most. The technology isn't the crime. Plenty of these silent-video accounts openly label themselves as A.I. characters — performance, not fraud. The exact same tool becomes a scam the moment someone hides what it is and uses it to take from you. The threat was never the code. It's the intent to deceive.

So here's the whole thing in plain terms. Scammers are building fake people that look completely real, and using them to make you fall first. The videos stay silent to dodge the tools that used to catch them. And the losses doubled in one year. The old test — "let me search this photo" — is fading. The new test is simpler and harder. Can they show up? Whether you investigate these cases or just check your messages at night, the safest question now isn't "is this photo real?" It's "will this person ever be in the same room as me?" The full story's in the description if you want the deep dive.

That Hot Stranger Sliding Into Your DMs? Probably 40,000 Lines of Code.