New anonymized data from data from Findmykids, a family safety app, shows that 56.3% of US children aged 9 to 17 had at least one AI app on their phone in September 2026, up from 41.6% a year earlier. The average number of AI apps per child grew 54%. This sharp rise signals that AI is no longer a niche tool for tech-savvy teens but a mainstream fixture of childhood, with implications for how kids learn, socialize, and spend their free time.
Perhaps most striking is the rapid adoption among the youngest users. Among 9-10 year-olds, AI app ownership jumped from 37% to 53%, approaching the 58% rate among teens. This is despite the fact that popular chatbots like ChatGPT are not intended for children under 13. The finding raises questions about whether age restrictions are being enforced and whether parents are fully aware of what their children are accessing.
Usage patterns reveal a critical split. Chatbots such as ChatGPT function like a search box: daily users grew 71%, but time spent per session remained flat. In contrast, AI companions—characters with a name, personality, and memory—are installed on only 8.7% of phones but command intense engagement. Weekend use peaks at 103 minutes a day, and the average companion user spends 93 minutes daily, compared with just 18 minutes for the average chatbot user. As the Findmykids Analytics Team noted, “That gap, not the install count, is what parents should watch.” The immersive nature of AI companions means they can hold a child's attention for hours, potentially displacing other activities and creating emotional dependencies.
The report also highlights that restrictions on one app do not eliminate the category. After Character.AI ended open-ended chats for minors, its daily users more than halved, but Talkie grew from 3.3% to 4.9% of phones. This migration suggests that kids will seek out alternatives when one platform closes its doors, making it difficult for parents and regulators to keep up.
Safety concerns are mounting. Nearly half of teens using AI chatbots report at least one harmful experience, according to a national Cyberbullying Research Center survey. Sameer Hinduja, the center's co-director, writes that “chatbots have moved past novelty status for many young people.” Harmful experiences can range from exposure to inappropriate content to emotional manipulation, and the risk may be even higher for younger children who lack the critical thinking skills to navigate such interactions.
For parents, the challenge is twofold: knowing which apps are installed and understanding how much time children spend on them. Findmykids helps by showing parents which apps, including new AI apps, a child has installed and the child's daily screen time. The full study is available on the Findmykids blog. However, the data reflects Findmykids users only, not all US children, so the true national picture may differ.
The broader implication is that AI is becoming an integral part of childhood, often without adequate guardrails. As adoption continues to climb, parents, educators, and policymakers must grapple with how to protect children from potential harms while allowing them to benefit from the technology. The engagement gap between chatbots and companions is a warning sign: it's not just about whether kids use AI, but how deeply they are drawn in.
