Table of Contents
One in ten teens now knows someone who has been targeted by a deepfake, an AI-generated image or video built to look real. That number didn’t exist five years ago, and it’s landing right as kids head back into hallways, group chats, and the apps where this kind of bullying spreads fastest. Cyberbullying used to mean a mean comment or a group chat someone got left out of. This year, child safety experts say the tools have changed, and most parents haven’t caught up.
What’s Actually Happening
Rania Mankarious, CEO of Crime Stoppers of Houston, has spent the run-up to this school year warning parents that the shape of cyberbullying is shifting fast. “You want to be talking to your kids about what cyberbullying looks like, and you as a parent need to know it’s changing,” she said. AI tools now make it simple to build a fake social media account, alter a photo, or generate new content using a real child’s face and name, work that once took technical skill and now takes an app and a few minutes.
The scale backs up her warning. A RAND survey of school principals found that 13 percent had dealt with a bullying incident involving an AI-generated deepfake over the past two school years, with the number climbing to 22 percent among high school principals and 20 percent among middle school principals. The National Center for Missing and Exploited Children reported that AI-generated child sexual abuse material sent to its tip line nearly doubled in the first six months of last year alone.
None of this requires a child to be a target of anything sexual to cause harm. A fake account built in a classmate’s name, a manipulated photo dropped into a group chat, an AI-written message sent to look like it came from someone else. All of it can spread through a school in a single afternoon, and a parent watching their child’s actual phone use might never see any of it happen. So much of it moves through group chats and disappearing messages rather than a public post.
Cyberbullying overall was already climbing before AI became part of the mix. The Cyberbullying Research Center’s most recent national survey found 58 percent of US teens have experienced cyberbullying at some point, and 33 percent were bullied online in just the past 30 days, the highest short-term rate the center has recorded. Girls report it more than boys, and LGBTQ+ teens report it at meaningfully higher rates than their peers, according to CDC data.
How to Spot Something Isn’t Real
Kids don’t need to become forensic analysts, but a few habits help. Reverse image searches (right-click, search the image, or use a phone’s built-in image lookup) can surface where a photo actually originated. Odd details are still common giveaways in AI-generated images: hands with too many or too few fingers, text in the background that looks garbled, lighting that doesn’t quite match between a face and its background, or an account with almost no post history suddenly posting something dramatic.
Encourage a simple habit before reacting to anything upsetting online: pause, and ask a trusted adult or a close friend to look at it too, before sharing it further or believing it outright. A second set of eyes catches a lot, and slowing down for even thirty seconds keeps a fake from spreading as fast as it would otherwise. This habit is just as useful for parents scrolling their own feeds. Adults share unverified viral content just as often as teens do, and a parent who models a quick pause before sharing something shocking teaches the same skill without ever needing to lecture about it directly.
What Experts Say Parents Should Actually Do
The advice from child safety groups has shifted from “monitor everything” toward something more realistic. Mankarious’s core recommendation: talk to your child about this before there’s a problem, not after. “They need to be talking to kids about why it’s dangerous, why it breaks a moral code of conduct, how it can get you in trouble,” she said, pointing out that plenty of kids assume a fake image is harmless, reasoning that nobody actually got hurt.
Common Sense Media’s guidance for these conversations leans on curiosity rather than a lecture. Instead of asking “do you know about AI,” they recommend something more like, “Have you ever seen something online that looked real but turned out to be fake?” Teens open up more when the conversation feels like a two-way exchange rather than a warning. Several short conversations work better than one long one, and expecting a brief, slightly awkward response is normal, not a sign the conversation failed.
If a child does come forward, experts are consistent on one point: don’t react with blame or an immediate punishment. A scared kid who gets grounded for telling a parent about a fake account or an altered photo learns not to tell them next time. The recommended first move is to listen, then preserve the evidence, screenshots, usernames, dates, and any messages, before looping in the school if other students are involved.
What This Means for Your Family
You don’t need to know every app your child uses. You do need to know roughly where they spend time online, and your child needs to know they can bring you a problem without immediate consequences landing on them. That single piece of trust does more than any monitoring app.
For younger kids (elementary age), the conversation is simpler: not everything online is real, and if something upsetting shows up on a screen, tell an adult right away, no exceptions and no trouble for asking. For tweens and younger teens, add the specifics: fake accounts, altered photos, and AI tools exist, using them to hurt someone is never acceptable, and being on the receiving end of one is never something to hide out of embarrassment. For older teens, bring in the legal reality directly. A federal law known as the TAKE IT DOWN Act now makes it a crime to publish nonconsensual intimate images, including AI-generated ones, with penalties of up to three years in prison for images involving a minor. As of this year, 46 states have also passed their own laws criminalizing AI-generated child sexual abuse material, regardless of whether a real photo of the child was used to make it. Florida classifies knowing possession of an AI-generated sexual image of an identifiable person as a felony, and Colorado allows up to three years of imprisonment for sharing a deepfake of a minor. In plain terms: a classmate creating one of these images isn’t just cruel, it can mean a felony charge.
A device curfew and a quick weekly check-in on privacy settings sound basic, but they’re still the most effective baseline experts recommend, not for catching everything, but for keeping the conversation about online life a normal, ongoing part of family life instead of something that only comes up after a crisis.
How Schools Are Responding
Schools are behind the pace of the technology, and most administrators know it. A 2023 Texas Education Poll found 74 percent of Texans rated cyberbullying a moderate or large risk to public school students, the top-ranked concern in the survey, well before AI-generated content became part of the equation in a serious way. Texas law already requires districts to address cyberbullying that affects students even when it happens off campus, and several other states have similar rules on the books. In practice, that means a fake account or an altered photo created outside school hours can still trigger a school response if it disrupts a student’s ability to learn or feel safe on campus.
Districts vary widely in how well they enforce this. Some have added reporting tools inside the platforms kids already use for schoolwork; others still rely on a student or parent walking into the front office with a printed screenshot. If your child’s school doesn’t have a clear, visible process for reporting this kind of incident, asking the front office directly is a reasonable place to start, and it often prompts the district to point families toward a policy that already exists but isn’t well advertised.
The Bigger Trend
What’s landing on parents this year isn’t really new territory, kids have always tested boundaries around image and reputation. What’s different is the speed and realism AI adds to something that used to take real effort to fake convincingly. A rumor used to spread by word of mouth. A doctored photo used to take some skill. Now both move at the speed of a group chat, built by a tool that took someone’s fifteen-year-old classmate a few minutes to use. The families having better outcomes this year aren’t the ones with the strictest rules. They’re the ones who started the conversation before their kid needed to come to them with a problem. That single shift, treating this as an ordinary topic instead of a rare, heavy talk, is what turns a scary headline into a manageable part of raising a kid in a school where a phone is basically standard equipment.