
Deepfakes have gone from impressive curiosities to background noise. Most teens see several a week without realising it. Here are three patterns we've spotted recently and the small checks that catch them.
Pattern one: the celebrity endorsement. A famous face promoting a product they would obviously never touch. The tells are in the eyes (they don't quite track) and the cadence of speech (slightly too even). The fix: search the celebrity's actual social account. If they aren't talking about it there, they aren't talking about it.
Pattern two: the 'leaked' clip of a politician saying something outrageous. Check the lighting on the face versus the background. Deepfakes are still bad at matching colour temperature. And reverse-image search a frame — the original source will usually surface fast.
Pattern three: the AI-generated photo of a 'real event' that never happened. Look at hands, ears and small text in the background — signs and posters often contain garbled letters. Then check whether any actual news outlet is reporting it. If only random accounts are, it's almost certainly fake.
None of these checks takes longer than thirty seconds. Make them a habit and you've already given your child a meaningful edge.
Continue reading
What the New York Times got right about kids and AI
A NYT opinion piece argues parents — not schools — are the front line of AI literacy. We agree. Here's what to actually do this week.
Welcome to FITinAI: why we built it
Most AI advice for families is either techno-utopian or doom-laden. We wanted a third option: calm, structured, honest.