Teaching Kids to Spot Deepfakes and AI Misinformation (Before They Share It)
A 12-year-old at a school I know sent a video to her group chat of a famous singer “admitting” they use AI to write all their songs. Within an hour the video had spread to half the grade. Kids were arguing about it at lunch. Two friends got into a real fight over it. The next day someone pointed out the video was a deepfake. The singer never said any of it. But by then the damage was done, the arguments had already happened, and a handful of kids still didn’t believe the correction.
This is the kind of moment most parents aren’t prepared for, because the technology that makes it possible got good fast. Three years ago, deepfakes were obvious. Today, the best ones are indistinguishable from real footage. Voices can be cloned from a few seconds of audio. Images can be generated that look like photographs. Articles can be written that sound exactly like a real news outlet.
Your kid is going to encounter this content. They’re probably already encountering it. The question is whether they have the skills to slow down before they share, and whether they know what to do when they’re not sure what’s real.
What kids are actually up against
The first thing worth understanding is that AI didn’t create the misinformation problem, it just made the volume go up by several orders of magnitude. Anyone can now generate a convincing fake article, fake image, or fake video in under a minute. The cost of producing lies has gone to zero, and the quality has gone way up.
A study out of MIT found that false news on Twitter spreads about six times faster than true news. Fake stories tend to be more surprising, more emotional, and more shareable than the truth, and algorithms reward engagement over accuracy every time. Outrage drives clicks, and lies are often more outrageous than reality. Your kid’s feed is calibrated to surface the most engaging content, which is frequently the least true.
That’s the environment they’re scrolling through. Most kids assume that if something looks real, it probably is. Most kids assume that if their friend sent it, it’s been verified. Most kids assume that one source is enough. None of those assumptions hold up anymore.
The method I teach: SIFT
The technique that holds up best for kids is called SIFT. It was developed by digital literacy researcher Mike Caulfield for college students, but the framework works just as well for middle schoolers because the steps are short and you can actually remember them under the pressure of “should I share this?”
S is for Stop. Before you do anything, pause. Don’t share it yet. Don’t comment. Don’t react. The whole misinformation machine depends on people sharing fast and thinking later. The single most powerful thing your kid can learn is the pause.
I is for Investigate the source. Who actually made this? Is it a news organization with a real reputation, or is it a website nobody has heard of? Is it from a verified account or an anonymous one? A 30-second check on the source resolves most fake content immediately, because most of it comes from sources that wouldn’t pass even a quick sniff test.
F is for Find better coverage. If the story is real, other reliable sources will be reporting it. If nobody else is covering it, that’s a red flag. If the only outlets covering it are sites with names like “TruthBuzz” or “RealNewsToday,” that’s a much bigger red flag. Real news gets reported by multiple real outlets.
T is for Trace claims to the original. Where did this image first appear? Where did this quote first come from? A lot of misinformation is real content stripped of its context, or fake content presented as real. Tracing it back to the original source usually clears it up fast.
You can teach the whole thing in ten minutes. Your kid won’t always use it, but if you’ve drilled the four letters enough times, the “Stop” part will kick in at the right moment, and that alone prevents most of the damage.
Walking through a real example with your kid
Pick something that’s actually trending. An image they saw on TikTok, a video their friend sent, a headline from their feed. Sit down together and walk through SIFT on the actual piece of content. This is way more useful than a generic explanation, because they’ll see how fast the framework works on something real.
Start with Stop. Just notice that they’re tempted to react. Ask them what they would have done with this two minutes ago. Most kids will admit they would have shared it or believed it without checking. That moment of self-awareness is the lesson.
Then Investigate the source. Open the account or website that posted it. How old is the account? How many followers? Is it verified? Does it look like a real news outlet or like something thrown together yesterday? Show your kid how to actually check, because most kids have never thought to click on the profile of the account that posted something.
Find better coverage by searching the key claim in a search engine. If the story is real, you’ll find it on real news sites. If it’s fake, you’ll find fact-checks debunking it, or you’ll find nothing at all. Both of those outcomes are useful information.
Trace it back. If it’s an image, do a reverse image search. If it’s a quote, find the original interview. Most fake content collapses the moment you try to verify it at the source, because the source either doesn’t exist or says something completely different from what’s being claimed.
The red flags worth memorizing
Beyond SIFT, there’s a short list of warning signs that should trigger automatic skepticism. Teach your kid to notice when content has any of these qualities, because the combination is almost always a tell.
It’s surprising or shocking in a way that feels designed to make you react. It’s posted by an account you’ve never heard of. The source is a website with a generic-sounding name and no clear ownership. There’s no author, no date, no original source linked. The headline doesn’t match the actual content of the article. It’s pushing you to share it right now. It’s trying to make you angry or scared at a specific group of people.
Any one of these can be innocent. Two or three together is a pattern. Four or more is almost certainly fake or manipulated. Once your kid has the list internalized, they’ll start spotting these in their own feed without having to think about it.
What to do when your kid has already shared something fake
This is the situation that brings most parents to this topic in the first place. Your kid shared something that turned out to be false. Now they’re embarrassed, or defensive, or trying to pretend they didn’t share it.
The instinct is to lecture. Resist it. The lecture is the thing that makes them shut down and refuse to take this seriously next time. What works better is a conversation that takes the mistake seriously without making them feel like an idiot for it.
Ask them what they thought when they first saw it. What made them believe it? What would they look for next time? Walk through SIFT on the post they shared, after the fact, so they can see what they would have caught if they’d paused. Then ask them what they want to do about it now. Do they want to post a correction? Send a follow-up to the group chat? Just let it die? There’s no single right answer, but the act of thinking through the cleanup is part of the learning.
The most important thing to communicate is that this is a skill, not a moral judgment. Adults share fake content too. Smart people share fake content. The kids who get good at this aren’t the ones who never made the mistake, they’re the ones who learned from it and slowed down the next time.
· · ·
When you want a structured way to teach this
If you’d rather not be the one explaining SIFT and deepfakes from scratch, Behind the Bot: A Kid’s Workbook to Use AI Safely, Think Critically & Stay in Control has a whole section on it. The workbook is for ages 10–14 and covers AI more broadly, but Section 4 (“Real or Fake?”) is dedicated to exactly this skill set.
What I like about teaching it through the workbook instead of a parent-led conversation is that the activities make the kid the detective. They work through scenarios where they have to decide if something is AI-generated or real and explain their reasoning. They build their own truth-finding toolkit. They read a story about a kid who shared a deepfake before checking and lost trust with her friends. They practice SIFT on six real-sounding claims with mixed sources.
By the time they finish the section, they’ve internalized the framework through doing it, not through being told about it. And the next time they see a suspicious video in a group chat, the pause kicks in on its own.
You can check it out on Amazon here.
· · ·
The kids growing up right now are the first generation that will never be able to fully trust what they see on a screen. That sounds dark, but it doesn’t have to be. The fix isn’t to keep them off screens or pretend the technology will go away. It’s to give them the tools to think critically about what they’re looking at, and the patience to slow down before they share. Those skills will matter for the rest of their lives. The earlier they build them, the better.