Your Kid's School Photo Is All a Classmate Needs to Make Deepfake Porn

Zero. That is the number of explicit photos a predator — or a classmate — needs to create a sexual deepfake of your child. Not one. Not a handful. Zero. They need a face. Just a face. The kind your kid posts every single day.

This week, multiple middle school students in South Korea were referred to authorities after they used photos of female classmates to create sexually explicit deepfake images. According to The Korea Herald, four middle schoolers were charged, with nine identified victims — all girls, all at school with the students who targeted them. No hacking. No stolen passwords. Just photos that were already out there, fed into an app, and turned into something horrifying in minutes.

South Korea is not an outlier. This is happening in Beverly Hills, California. In New Jersey. In Massachusetts. In Pennsylvania. It is happening in ordinary schools, in ordinary towns, to ordinary kids — and most parents have no idea what to do when it lands in their child's grade.

This Isn't a Stranger-Danger Story Anymore

We have spent years teaching kids about online strangers. Don't give your address to someone you've never met. Don't send photos to people you don't know. Good advice. Outdated framing.

The kids doing this — in most documented cases — are not strangers. They sit three rows away in math class. They were at the birthday party last year. They follow your kid on Instagram. The threat has moved from the unknown person in the dark to the known person in the hallway, and our safety scripts haven't caught up. This article is part of a series — start with Face Detection Before Identification How Facial Analysis Act.

For as little as $4.99, a teenager can upload a classmate's face photo to a "nudify" site — a tool that strips clothing from images using AI — and get back a sexually explicit fake image in seconds. No technical skill. No special equipment. Just pocket money and a photo that was already public. The Boston Globe documented this pattern across Massachusetts schools, finding that only a fraction of districts even have policies addressing deepfakes — let alone consequences for creating them.

Let that number breathe for a second. One in 25. Picture a typical classroom of 25 kids. UNICEF's global data is not describing a fringe problem in far-off places. It is describing a statistical near-certainty inside ordinary schools right now.

The Number That Should Keep You Up at Night

Reports of AI-generated child sexual abuse material (fake sexual content involving minors) sent to the National Center for Missing and Exploited Children jumped from 4,700 in 2023 to more than 400,000 in just the first half of 2025. Not a typo. Four hundred thousand in six months. That is not a trend line. That is a cliff.

RAINN — the nation's largest anti-sexual-violence organization — is unambiguous: AI-generated fake images of children are illegal under federal law in the United States, even when no real explicit photo ever existed. The harm to the child is real. The law treats it as real. The problem is that most schools and many police departments don't yet have a clear procedure for what to do when a parent shows up with a screenshot.

"When the image is fake but the harm is real, the legal system is struggling to keep pace — victims face reputational injury, emotional trauma, and social consequences that don't disappear when the image is taken down." — American Bar Association, Children's Rights Newsletter

That last part matters: when the image is taken down, the harm doesn't go away. The girl who sits in second period still has to sit in second period tomorrow. That is not a technology problem. That is a human one. Previously in this series: Your Face Is Your Passport Now And You Never Agreed To It.

What Schools Were Built For — and What They Weren't

Schools are extraordinary institutions when it comes to the problems they were designed to handle: physical fights, drug possession, bullying you can see. They have protocols. They have trained staff. They have forms.

They do not have a protocol for "student A used student B's yearbook photo to generate a fake sexual image and shared it in a group chat." Most administrators have never had to preserve digital evidence for a synthetic media investigation. Most school resource officers have never filed a report involving AI-generated content. Stanford HAI's policy analysis on AI-generated child sexual abuse material makes the institutional gap plain: schools built their child protection policies for a pre-AI world, and the policies haven't been rebuilt.

By the time most districts get around to updating their student conduct codes, another group of girls will have spent a school year trying to concentrate in class while knowing an image of them is circulating in group chats.

The Conversation You Actually Need to Have

Here is what "the talk" needs to include now. Not instead of everything else — in addition to it.

Your child's face is identity data (meaning it belongs to them the way a fingerprint or a signature does — it's uniquely theirs, and it can be misused). An innocent photo from a soccer game or a school event can be fed into an AI tool and turned into something sexual. That is not your child's fault, and it is not a reason to hide from the world. But it is a reason to treat face photos the way you treat passwords — with some thought about where they go and who can see them. Up next: Before Facial Recognition Names You It Has To Find You And T.

That is a different conversation than "don't post selfies." It's more specific. A public Instagram account where every photo shows a clear, full face is a different risk than a private one where photos are shared with actual friends. Group chats where someone can screenshot are a different risk than direct messages. These are real choices your kid can make — not vague warnings that make them feel surveilled.

If you have ever wondered whether someone online is really who they say they are — whether a photo is real, whether a profile is genuine — that is exactly the right instinct to have right now. Tools that help verify identity and detect synthetic media (AI-generated fakes) exist and are getting better fast. Knowing that the problem is being worked on is not the same as the problem being solved, but it's worth knowing you're not the only one paying attention to it.

The students referred to authorities in South Korea this week almost certainly thought of what they did as a prank. A joke. Something edgy they found online. Their victims — nine middle school girls — will carry what happened to them for a long time after those students have moved on. That gap between what the person who made the image experienced and what the person in the image experienced is the whole story. And right now, the only thing standing between your child and that gap is a $4.99 app and a photo they posted last Tuesday.

If your child told you tonight that a classmate had made a fake sexual image using someone's face, would you know what to do in the first hour? Not the first week — the first hour. That answer is worth knowing before you need it.