By: James Vance – SeaPRwire -Parents think they know what their kids do online. The numbers say otherwise. A new FOSI survey shows an 11-point gap on generative AI use alone. Kids report more activity across the board. This disconnect creates real risks in homes from California to Sydney. Tech moves fast. Awareness lags behind.
The Family Online Safety Institute released its fourth wave report in spring 2026. Ipsos surveyed over 4,000 parents and children aged 10-17 in the US and Australia. When asked about generative AI use in the past week, 27% of parents said yes. Children answered 38% for themselves. Similar gaps appeared elsewhere. Social media scrolling showed 46% of parents versus 54% of kids. Posting stood at 30% versus 38%. The pattern held for schoolwork too. Parents view it as safe. Kids still do more than reported. The perception divide runs deep and steady.
US children’s generative AI use has leveled off. It jumped 13 points from Wave One in fall 2024 to Wave Three in fall 2025. Wave Four shows 72% compared to 74% previously. Flat. Parents grew less optimistic about AI helping online safety. Their positive views dropped 10 points from 52% to 42% over the same waves. Trust in tech companies remains low. Only 27% of parents believe these firms effectively protect kids from harmful content. Children rate them slightly better at 41%. Family talks work better than anything else. Nine in ten kids say they can turn to parents if something online feels unsafe. Both sides agree parents bear primary responsibility for safety education. 92% of children and 91% of parents said so.
Rules beat software in most households. Families set household guidelines more often than installing parental controls. On smartphones, 68% have at least one rule. Only 49% use device controls. When controls exist, 81 to 86% of families say they function well. Countries differ on outside help. Australian families prove nearly three times more likely than American ones to expect government involvement in online safety education. 30% versus 12%. Teachers matter more down under too. 64% of Australians see a teacher role compared to 42% in the US.
I sat across from a fellow editor last month in New York. He laughed about his teenager’s “harmless” ChatGPT use for homework. Days later the kid admitted generating full essays and using image tools for fake project visuals. Stories like this repeat in parent groups everywhere. The FOSI data backs them up. Kids hide AI experiments. Parents underestimate exposure. The gap widens with newer tools. Generative AI sits at the center now. Its creative and deceptive potential demands closer watch.
Alanna Powers-O’Brien, Research Specialist at FOSI, noted the mix of agreement and division between parents and teens. Policymakers worldwide watch youth internet use closely. This report shows families navigating safety on their own terms. Tech giants like Apple, Google, TikTok, Roblox, Snap, Nintendo and Amazon sit on FOSI’s member list. Disney and TikTok supported the research. Low trust in companies persists despite their involvement. Parents prefer talking and rules over relying on corporate tools.
The data paints a practical picture. Conversation remains the top safety net. Build it early and keep it open. Check phone rules weekly instead of trusting default controls. Ask direct questions about AI tools. Don’t accept vague answers. Track trends across waves. US AI adoption stabilized but stays high at around 72%. Optimism fell. That signals caution. Australian families lean on institutions more. US ones stay home-focused. Adapt strategies to local expectations.
Parents cannot close the gap by guessing. They must act on evidence. Start with honest talks this week. Review one device together. Discuss a recent AI interaction. Make it routine. The numbers from over 4,000 families don’t lie. Kids do more online than admitted. Ignoring the 11-point AI blind spot invites trouble. Better to face it now with clear eyes and steady rules.
Author bio: James Vance, longtime senior commentator for international tech publications, analyzing digital culture shifts and family technology challenges.