Findmykids data shows that AI apps became significantly more common on U.S. children’s phones over the past year. In September 2026, 56.3% of children aged 9\u201317 in the dataset had at least one AI app installed, up from 41.6% a year earlier.<\/p>\n
But more AI apps did not necessarily mean more time spent using them. While chatbot adoption increased substantially, daily usage remained broadly stable. AI roleplay apps showed a different pattern: they reached fewer children but generated much longer usage among those who had them.<\/p>\n
AI apps became considerably more common on children’s phones between September 2025 and September 2026.<\/p>\n
The share of children with at least one AI app installed increased from 41.6% to 56.3%<\/strong> \u2014 an increase of 14.7% in one year<\/strong>. In other words, more than half of the children in the dataset had at least one AI app on their phones by September 2026. This average includes all children in the dataset, including those without any AI apps installed.<\/p>\n Most children with AI apps still had only one installed, but the distribution shifted toward multiple AI tools.<\/p>\n In September 2025, 70.4% of children with at least one AI app had only one installed<\/strong>. By September 2026, that share had fallen to 60.4%<\/strong>. Taken together, the figures show that AI use became not only more widespread across children\u2019s devices but also more varied among children who already had access to AI apps.<\/p>\n Not all types of AI apps are equally common.<\/p>\n General-purpose chatbots were by far the most widespread category: 53% of children had at least one chatbot app<\/strong> on their phones. The chatbot category includes ChatGPT, Gemini, Claude, Copilot, Perplexity, Grok, and DeepSeek.<\/p>\n AI roleplay apps were much less common, with 8.7% of children<\/strong> having at least one app in this category. This group includes Talkie, PolyBuzz, Character.AI, and Chai.<\/p>\n Study-focused AI apps were the least common of the three categories, found on the phones of 1.8% of children<\/strong>. These include Grammarly, QuillBot, Photomath, Mathway, Symbolab, Question.AI, Brainly, Socratic, and Gauth.<\/p>\n Chatbots also showed the strongest growth. The share of children with at least one chatbot app increased by approximately 16%<\/strong> between September 2025 and September 2026. Roleplay AI increased by around 2%<\/strong>, while dedicated study-focused AI showed little growth.<\/p>\n Despite growing access to AI overall, dedicated study-focused AI apps remained uncommon on children\u2019s devices.<\/p>\n Only 1.8% of children had a study-focused AI app<\/strong>, compared with 53% who had at least one general-purpose chatbot<\/strong>.<\/p>\n Study-focused AI was also the only category in this analysis to show a decline between the comparison points available in the data. The share of children with at least one dedicated study app fell from 2.1% in May 2025 to 1.8% in September 2026<\/strong>.<\/p>\n The Findmykids data cannot establish why dedicated study apps remain less common. One possible explanation is that children may use general-purpose chatbots for tasks that could otherwise be handled by dedicated study tools. However, the available data cannot confirm this hypothesis, and further research would be needed to test it.<\/p>\n AI apps became more common across every age group analyzed. The share of children with at least one AI app increased year over year in all four groups:<\/p>\n By September 2026, more than half of children in every age group analyzed had at least one AI app installed<\/strong>.<\/p>\n Older children were somewhat more likely to have AI apps, but the largest year-over-year increase occurred among the youngest children in the dataset: +16.4 percentage points among 9\u201310-year-olds<\/strong>, compared with +12.2 percentage points among 15\u201317-year-olds<\/strong>.<\/p>\n The rapid growth in AI app installations was not accompanied by a comparable increase in time spent using chatbots.<\/p>\n Average daily chatbot use changed only slightly, from 19 to 18 minutes per day<\/strong>, while median daily chatbot use remained at approximately 5 minutes per day<\/strong>.<\/p>\n This highlights an important distinction between AI adoption and AI engagement<\/strong>. More children had AI apps installed, and children increasingly had multiple AI tools available on their devices, but the available usage data did not show a corresponding increase in average time spent using chatbots.<\/p>\n Data note:<\/strong> Gemini is not included in the available usage-time analysis despite its growth in installations. Usage trends should therefore be interpreted with caution.<\/p>\n AI roleplay apps reach far fewer children than general-purpose chatbots. In September 2026, 8.7% of children had at least one roleplay AI app<\/strong> installed.<\/p>\n At the same time, children who used these apps spent considerably more time in them than chatbot users.<\/p>\n Average daily roleplay AI use decreased from 105 to 93 minutes<\/strong>, a decline of approximately 11%<\/strong> year over year. Median daily usage also fell, from 54 to 48 minutes<\/strong>, or about 12%<\/strong>.<\/p>\n Despite these declines, roleplay AI remained associated with substantially longer daily usage than general-purpose chatbots. In September 2026, average daily usage was:<\/p>\n The data therefore points to a different engagement pattern: roleplay AI apps reach a smaller share of children, but children who use them spend considerably more time in them on average<\/strong>.<\/p>\n Usage patterns also differed between weekdays and weekends.<\/p>\n \u2022 For general-purpose chatbots, average daily use changed only slightly, from 18 minutes on weekdays to 17 minutes on weekends<\/strong>. The data therefore shows a distinct weekend pattern for roleplay AI: children who use these apps spend substantially more time in them than chatbot users, and their average usage increases further on weekends.<\/p>\n This study is based on anonymized data from the Findmykids app<\/a>, a family safety app that combines parental controls and GPS tracking to help families protect the online and offline safety of children and other loved ones.<\/p>\n Four comparable periods in 2025 and 2026 were analyzed.<\/p>\n The headline year-over-year findings in this research primarily use the September 2025 vs. September 2026 comparison.<\/p>\n For this analysis, AI applications were grouped into three categories.<\/p>\n ChatGPT \u00b7 Gemini \u00b7 Claude \u00b7 Copilot \u00b7 Perplexity \u00b7 Grok \u00b7 DeepSeek<\/p>\n Talkie \u00b7 PolyBuzz \u00b7 Character.AI \u00b7 Chai<\/p>\n Grammarly \u00b7 QuillBot \u00b7 Photomath \u00b7 Mathway \u00b7 Symbolab \u00b7 Question.AI \u00b7 Brainly \u00b7 Socratic \u00b7 Gauth<\/p>\n The analysis looked at the share of children with at least one AI app installed<\/strong>, the average number of AI apps installed per child<\/strong>, the distribution of children with one, two, or three or more AI apps<\/strong>, and adoption across the three AI app categories.<\/p>\n Where usage data was available, the analysis also considered average and median daily usage<\/strong>, differences between weekdays and weekends<\/strong>, and patterns across age groups<\/strong>.<\/p>\n Need additional statistics for your story? We may be able to provide another age group, GEO, time period, or related aggregated Findmykids data.<\/p>\n
\n\u2800
\nThe number of AI apps found on children\u2019s devices increased as well. The average number of AI apps per child rose from 0.58 to 0.89, an increase of approximately 54%.<\/p>\nKids Are Increasingly Using More Than One AI App<\/h2>\n
\n\u2800
\nAt the same time, having multiple AI apps became more common. The share of children with two AI apps increased from 22.7% to 27.2%<\/strong>, while the share with three or more rose from 5.5% to 8.8%<\/strong>.<\/p>\nChatbots Dominate Kids’ AI App Adoption<\/h2>\n
Dedicated AI Study Apps Remain Rare<\/h2>\n
More Than Half of Kids in Every Age Group Now Have an AI App<\/h2>\n
\n
More AI Apps Doesn’t Mean More Time Spent Using Them<\/h2>\n
Roleplay AI Has a Smaller Audience \u2014 but Much Longer Usage<\/h2>\n
\n
Roleplay AI Use Is Higher on Weekends<\/h2>\n
\n\u2022 Study-focused AI showed a similar pattern, with usage falling from 15 to 14 minutes per day<\/strong>.
\n\u2022 Roleplay AI showed a more pronounced difference. Average daily use increased from 89 minutes on weekdays to 103 minutes on weekends<\/strong> \u2014 approximately 14 minutes more per day<\/strong>.<\/p>\nMethodology<\/h2>\n
Population<\/h3>\n
\n
Analysis Period<\/h3>\n
\n
What Counts as an AI App?<\/h3>\n
Chatbots<\/h4>\n
AI Roleplay<\/h4>\n
Study-focused AI<\/h4>\n
What We Measured<\/h3>\n
Need More Data?<\/h2>\n
Request Data<\/a><\/h3>\n\t\t