A couple of months ago, one of my students showed me a video on his phone. In it, Jake Paul, the YouTuber turned professional boxer was fly fishing on a pristine river. Then, he turned to the camera, looked directly into it, and said happy birthday to my student by name.
I did a double take. Then I realized it never happened. A classmate had built it in minutes using Sora, OpenAI's video tool. This AI tool generates high-definition, one-minute videos from a simple text prompt. No film crew. No editing suite. Just a phone and a few typed sentences.
The video was harmless. But I could not stop thinking: what if it wasn’t?
A New Kind of Media Literacy
Every generation of parents has had to teach their kids something new about media. Boomers explained TV advertising. Remember the house hippo? Millennials taught their children to fact-check viral posts. Gen Z learned to spot the staged “authenticity” of influencer culture. Now it is our turn.
Gen Alpha (children born after 2010) are the first generation growing up where artificial intelligence can produce hyper-realistic images and videos from a single text prompt. What makes this different is not just the technology, it is that AI removes the visual cues children have always relied on to separate real from fake. A talking animal signals fantasy. A blurry photo signals doubt. But a deepfake mimics natural lighting, authentic textures, and lifelike movement, the very things that signal “this is real.”
Photo caption: AI generated image of a cat in a box created in Google Gemini
Experts project that 10 million deepfakes will be shared in 2026, up from just 500,000 in 2023. And here is the sobering part: even adults detect them correctly only about 50% of the time, barely better than a coin flip.
Why This Matters in the Classroom and at Home
The stakes go beyond “fake news.” AI-generated imagery is already showing up in schoolyards. In several countries, police have issued warnings about students using “nudify” apps to create non-consensual images of their peers, images that cause real psychological harm, including anxiety and trauma. Many students do not realize this can be a criminal offense. Many parents do not know it is happening. In fact, only 37% of parents whose children use AI tools are even aware their child is using them at all.
This is not a reason to panic. It is a reason to talk.
What We Can Do: Four Practical Starting Points
1. Start the conversation before they encounter it. Ask your child: “Have you ever seen a photo or video that turned out to be fake?” Children who talk regularly about AI with a trusted adult are significantly more likely to pause and verify before sharing or believing what they see.
2. Teach them to look for the glitches. AI still leaves fingerprints. Encourage kids to notice unnatural shadows, hair that blurs at the edges, eyes that do not quite blink right, or hands with too many fingers. Making it a game: “spot what is wrong with this image”: builds the habit without the anxiety. A great place to practice: Which Face Is Real, a free site with a tutorial on spotting AI faces and a game where kids (and adults) test themselves with real vs. generated images.
3. Create a “reality check” habit. Before reacting to a striking image or video, model the question: “How do I know this is real?” Tools like Google Reverse Image Search take seconds and can reveal whether an image has been circulating in other contexts.
4. Ask your school about their AI policy. Schools should explicitly address deepfakes in their codes of conduct, not just general cyberbullying rules. If yours has not, this is a productive conversation to start with administration.
The Bigger Picture
Our children are not the first generation to inherit a media landscape more complicated than the one their parents grew up in. It surely will not be the last. What we give them is not immunity from confusion, but the tools to ask better questions.
Seeing is no longer automatically believing. Teaching kids to know the difference is one of the most important things we can do right now.