Creating nude deepfakes is rampant among young people — and AI safety education is lacking

The Classroom Silence on AI Safety Leaves Teens Exposed to a Growing Digital Threat

Healfromzero.com – Walk into most American middle or high school classrooms this academic year and you will find lessons on algebra, biology, and American history. What you will almost certainly not find is a structured conversation about how to navigate artificial intelligence without putting yourself at risk. A fresh survey conducted by Common Sense Media, a San Francisco–based nonprofit focused on children’s media and education, reveals that only three in ten teenagers aged thirteen to seventeen have ever had a teacher discuss safe AI use. The remaining seventy percent are essentially on their own, figuring out how to interact with chatbots, generative image tools, and voice assistants without formal guidance.

The gap matters because AI has already become woven into daily adolescent life. The same survey found that seventy percent of teens now rely on AI tools to complete homework assignments. Parents have long intuited that their children need coaching on how to ensure AI supplements rather than replaces genuine learning. But the conversation most families have had — or haven’t had — tends to orbit a narrow question: is the kid cheating? That framing, while understandable, misses a far more urgent danger unfolding in the background.

The Deepfake Problem No One Is Teaching Kids to Handle

Generative AI has made it trivially easy to produce photorealistic nude images of real people. These so-called “nude deepfakes” are AI-generated pictures or short videos that either depict someone in a pornographic scenario or digitally strip away their clothing. They are not hypothetical. They are proliferating among teenagers at a pace that has alarmed digital-forensics researchers.

“There is a fundamental underestimation and probably adult misconception about how often and how familiar adolescents are with using Gen AI,” said Dr. Chad Steel, an adjunct professor of digital forensics at George Mason University in Fairfax, Virginia, who was not involved in the Common Sense Media study.

Steel, who studies how digital evidence is created and manipulated, described the prevalence of adolescent-created nude deepfakes as behavior that has crossed into near-normalcy for the age group. Because the phenomenon is so recent, long-term psychological effects on victims remain unknown. What researchers do know, however, is that individuals whose intimate images circulate online can be tracked, contacted, or even stalked years later by people who have accessed that content. Predators have been known to threaten to forward such images to a victim’s parents or close friends unless monetary or other demands are met.

The downstream consequences extend well beyond embarrassment. As noted in prior public-health commentary, once intimate images of a person exist on the internet, they can complicate dating, hinder job prospects, and elevate the risk of sexual assault, clinical depression, and suicide. For a teenager whose social identity is still forming, that calculus is especially destabilizing.

Why Schools Have Not Caught Up

Michael Robb, head of research at Common Sense Media and coauthor of the survey, acknowledged that the absence of classroom instruction on AI safety is not a mystery. Most curricula were built before generative tools became ubiquitous, and teachers have not yet received standardized training on how to integrate digital-safety lessons around AI.

“For the most part, they’re just not happening yet,” Robb said of the conversations about AI literacy — including how the technology works, how to verify whether a chatbot’s answer is accurate, and how to protect oneself online.

Robb also flagged two methodological caveats. First, students may simply not recognize which parts of their coursework count as “AI instruction,” so the true exposure rate could be marginally higher than thirty percent. Second, even for the minority who do receive some guidance, the study cannot measure whether that instruction is effective. In other words, the data describes a vacuum, not a solution.

What Parents Can Do Starting Tonight

Until school systems build out formal curricula, the burden of conversation falls on households. Several concrete steps emerge from the research and from Steel’s forensic expertise:

Explain permanence. Many adolescents are comfortable with fashion apps that let them “try on” garments by draping a virtual outfit over an avatar. That familiarity can blur the line between a playful digital experiment and the creation of a nude image of a real person. Steel emphasized that teenagers “certainly don’t understand how to protect themselves, and that these images may be permanent, and may be out there forever.” Once a sexualized image is uploaded, anyone can save it to a private archive or personal device, making complete removal from the public sphere effectively impossible.

Set a clear rule. Talk with your child about why they should never create, share, or even view nude deepfakes. Frame it not as a moral judgment but as a practical risk-management decision, the same way you would discuss locking doors or checking a stranger’s identity before answering the phone.

Restrict photo exposure. Teach younger children and teens to post ordinary selfies and group photos only to private accounts where every follower is a known, trusted person. The logic is straightforward: once a clear image of a face is publicly accessible, anyone with a generative-AI tool can composite that face onto a nude body. Limiting stranger access to your child’s photos is the single most effective structural safeguard available.

Normalize the conversation early. These discussions do not need to wait until a child is sixteen or seventeen. Introducing the vocabulary — deepfake, generative AI, digital permanence — at younger ages builds a framework that makes later, more specific conversations easier and less awkward.

A Broader Implication

The thirty-percent figure from the Common Sense Media survey is not merely a statistic about curriculum gaps. It signals that an entire generation is entering adulthood with no institutional scaffolding for one of the most consequential technologies of the twenty-first century. The homework question — “are you cheating?” — is manageable. The question that follows, “do you know how to keep your own image, identity, and dignity intact in a world where anyone can generate a convincing nude of you in seconds?” — is the one that currently has no classroom answer. Until educators, policymakers, and parents collectively fill that silence, the risk sits squarely on families who may not yet know the threat exists.