According to "European children's rights in the age of generative AI" (LSE, 5Rights Foundation and EU Kids Online, October 2026), 72% of children aged 9–16 across 17 European countries used generative AI in the previous month, rising from 45% of 9–10-year-olds to 86% of 15–16-year-olds. Teens interviewed value it for learning but worry about misinformation, privacy, deepfakes and overreliance, and say they get little guidance from adults.
WhatsApp has launched optional parental controls for teens aged 13 to 17. Using a 6-digit parent PIN, parents can manage profile privacy, get alerts when their teen joins or leaves groups, and switch Meta AI, the app's built-in AI assistant, from its default 13+ setting to a stricter "Limited" mode. Parents cannot read chats or listen to calls, and the feature is rolling out gradually, so it may not be in your country yet.
Roblox Build is an AI tool inside the Roblox app that turns a written description into a playable game, no coding required. Roblox said on November 13, 2026, that it’s live in New Zealand, Singapore, Serbia, Thailand and Canada, coming to the U.S. later this year and worldwide early next year. Users must be 9 or older and pass an age check. It’s free to start, but extra AI usage costs Robux, so parents should set spending limits.
Character.AI, the character-chatbot app, is in Chapter 11 bankruptcy and plans to auction its assets on December 1, 2026. On November 20, a coalition of 40 state attorneys general announced that the court ordered an independent Consumer Privacy Ombudsman to review any sale involving user accounts, chats and personal data, including children’s. If your teen has an account, now is a good time to review it together and, if you choose, export and delete the data from account settings.
According to the Family Online Safety Institute's "Coming of Age" study, released November 19, 2026, 51% of U.S. teens aged 13-17 who use AI chatbots use them daily, and 46% say it feels like talking to a human. The upside: trust and conversation are already happening at home (90% of parents have talked about chatbot safety). The gap: parents and teens disagree on who is responsible for keeping teens safe, so it's worth spelling that out together.
A Gallup and Boston Children's Hospital Digital Wellness Lab study released in November 2026 surveyed 15,925 young people aged 15-22 in 16 countries. About a third have used AI as a social companion or friend, rising to half among daily AI users. The headline for families: in-person friendships and strong relationships with parents are most closely tied to wellbeing, while using AI for companionship is linked to higher loneliness, though the study can't show cause and effect.
In November 2026, Sesame Workshop, the nonprofit behind Sesame Street, released its "AI ABCs," five design principles for AI aimed at young children. The most important, says CEO Sherrie Westin, is that AI should support caregivers and never replace them, which is why Sesame won't make AI-powered Elmo toys. Parents can use the principles as a checklist before picking an AI app or toy.
On November 17, 2026, Oregon Attorney General Dan Rayfield filed a $2.035 billion legal action against OpenAI, alleging the company misled the public about how safe ChatGPT is for children and teens. These are allegations a court has yet to decide, but the state's advice applies to every family: parental controls and age checks don't catch everything, so keep talking with your kids.
On Nov. 16, 2026, OpenAI announced new measures against child sexual abuse material (CSAM): a free image classifier for vetted organizations that flags likely abuse images, including AI-generated ones; ChatGPT warnings that interrupt attempts to find such material, delete the conversation and pause access for 24 hours; guidance for developers; and tools for child-safety investigators. It's a meaningful step for platforms, but it isn't a parental control. Families still need open conversations and to know where to report, such as NCMEC's Take It Down and CyberTipline in the U.S.
Most U.S. parents are wary of AI in the classroom, according to an Institute for Family Studies (IFS) survey published Nov. 12, 2026. YouGov polled 2,000 parents of children ages 5 to 18: 56% want AI fully blocked in schools and another 26% want less of it than today. More than two in three want it introduced in high school or later, or never, and 77% expect it to hurt learning. Yet only 55% feel they can shape how AI is used in their child's classroom. The takeaway for families: get informed and get involved before decisions are made.
AI Quests is a free, web-based game from Google Research and the Stanford Accelerator for Learning, designed for students ages 11 to 14. Kids step into the role of researchers: they collect and clean data, train an AI model and test it, with no coding. On Oct. 26, 2026, Google launched a third quest, Polar Peaks, about mapping neurons in the brain, plus new teacher training modules. AI Quests requires no account, collects no personal information and uses no generative AI inside the game. It is built as a classroom resource with free teacher guides.
Yes, parents' habits and messages matter a lot. A RAND report published Nov. 12, 2026, based on nationally representative surveys of about 1,000 U.S. parents and 1,800 teens ages 12 to 18, found that teen AI use for homework rose from 40% to 49% between 2025 and 2026. Apart from age, the factor most closely tied to that use was whether teens felt their parents encouraged it. Teens of frequent AI-using parents used it more (63% vs. 40%), and an income gap opened up: 57% of teens in the highest-income families vs. 43% in the lowest. RAND urges parents to talk explicitly about the difference between AI doing the work and AI helping teens think.
SAFE-Teen, a study released on November 4, 2026, by researchers at Vals AI and Stanford University, tested 18 AI chatbots in 32,555 simulated conversations with teens ages 13 to 17 about mental health, health and relationships. With no teen-specific instructions, the chatbots failed 43.3% of applicable safety criteria; with a prompt stating the user's age and prioritizing safety, 24.9%. Even so, failures to respond to indirect suicide risk stayed at 52%, and safety slipped when teens pushed back or turned down outside help. The conversations were simulated and the paper is a preprint, but the takeaway for families is clear: a chatbot is no substitute for a trusted adult.
On November 10, 2026, the AFT and UFT teachers unions announced that Anthropic signed a legally binding agreement making their AI Safety and Privacy Standard for Schools enforceable for Claude for Teachers. Microsoft adopted the standard in September and has offered it to every U.S. school district since November 1. The standard sets 10 protections, including no AI training on student or educator data, no selling data or using it for ads, human review of high-stakes decisions, a ban on AI companions, breach notice within 72 hours and plain-language guides for families. It applies only in the U.S., but it works as a checklist for any school.
The Hour of AI is a free initiative from the nonprofit CodeAI (formerly Code.org) offering one-hour activities that help learners understand how artificial intelligence works, from pre-readers to adults, with no experience needed and some options that require no devices. New activities launch on November 16, 2026, and the big moment is Computer Science Education Week, December 7-11, 2026, though activities work any time of year. Kids under 13 can try most activities without an account, and CodeAI encourages parental oversight.
On September 17, 2026, the European Commission proposed the EU KIDS Act, which includes rules for AI chatbots and AI companions used by minors: no designs that simulate human relationships in ways likely to create emotional dependency, no carrying a child's earlier conversations into later ones by default, safety testing before launch and monitoring afterward, and access for under-13s only through parental controls. It is still a proposal that the European Parliament and Council must approve, and the Commission is collecting feedback from families and educators until December 3, 2026.
Yes. Starting December 2, 2026, the European Union's AI Act prohibits AI systems that create realistic sexual images, video or audio of an identifiable person without their consent (so-called "nudify" apps) and AI systems that generate child sexual abuse material. The rule applies in the EU, but the problem is global, and in schools it often starts with a classmate's photo. Beyond the law, what protects kids most is knowing they can tell a trusted adult without being blamed, and knowing about free tools like Take It Down that help get images removed.
On November 10, 2026, the Harvard Graduate School of Education (HGSE) launched Compass Collaborative and its free AI Guidance by Age tool. Parents and teachers pick their role, a developmental area (like literacy, creativity or emotions) and a child's age, then get a research-based take on how AI might help or hinder development at that stage, plus recommendations. It draws on more than 5,000 studies, doesn't rate specific apps and frames its guidance as hypotheses rather than final answers. The key question it asks: is this AI moving a child toward practicing the skill and connecting with people, or away from it?
In November 2026 Microsoft announced an upcoming option that lets schools lock Copilot Chat so students can only use Study and Learn, a tutoring agent that coaches with questions instead of doing the work; students can't switch out of it. Microsoft also introduced Microsoft Classroom, an app that brings together assignments, grades and the Study and Learn agent, with a preview planned for early 2027. Copilot Chat for students requires age 13+, and in K-12 schools it only works for 13- to 17-year-olds when a school admin turns it on.
On November 10, 2026, UNESCO published "AI Answers. You Decide," a free English-language companion to its Media and Information Literacy curriculum. Written mainly for teachers and teacher trainers, it offers activities that help children and teens critically evaluate AI-generated content, recognize bias and understand how algorithms shape what they see. The core message: AI can give answers, but deciding whether to trust, share or act on them is still up to people.
A Common Sense Media and SocialSphere survey released November 9, 2026, covering France, Germany and Italy, found that teens use AI more and are more optimistic than their parents: 52% of teens think it will help society, versus 28% of parents. Still, 79% of teens say young people need to learn to think critically without AI, only 36% always or often check whether its answers are accurate, and large majorities of both groups want stronger safeguards.
On November 10, 2026, the French presidency published a "Charter of commitment for safer artificial intelligence for children," backed by Anthropic, Black Forest Labs, Microsoft and OpenAI. The companies commit to principles such as safety and privacy by design, age-appropriate AI, tools for parents, age assurance that doesn't rely only on self-declared birthdays, and risk testing before and after launching AI products kids can use. It is a voluntary commitment, not a law.
On November 9, 2026, Meta announced it is expanding its AI age-estimation technology worldwide. If its systems suspect a user is a teen, even one who entered an adult birthday, they are moved into a Teen Account with built-in protections: a private account, messaging limits, content filters and time-management tools. The minimum age to have an account is 13.
According to GoStudent's Future of Education Report 2026 (a Censuswide survey of 3,001 parents, 3,001 children ages 8 to 16 and 300 teachers in six European countries), 68% of students who use AI for schoolwork admit handing in AI-generated content as their own. Only 20% have had a structured lesson on AI, and only half of parents feel confident supervising their child's AI use at home.
The German federation of consumer organizations (vzbv) found in a forsa survey that 93% of 14-to-25-year-olds used AI chatbots in the past year and 28% of those users have personal conversations with them. In its chat analysis, ChatGPT, Meta AI and Snapchat My AI came across as empathetic, human-like confidants and sometimes gave dangerous advice.
Common Sense Media rates AI chatbot toys an "unacceptable risk" and does not recommend them for children 5 and under, urging caution for ages 6 to 12. In its tests of three toys, 27% of outputs were not appropriate for kids, and the toys collect voice recordings, transcripts and behavioral data inside the home.
A Jed Foundation (JED) study released October 27, 2026, surveying more than 5,500 U.S. middle and high school students, found 58% use AI for schoolwork and 57% when they feel sad, stressed, lonely or overwhelmed. Using AI for companionship, relationship advice or while emotionally struggling was associated with more anxiety, depression, loneliness and history of suicidal thoughts. JED's core message: AI should never replace trusted adults.
On October 28, 2026, New York Governor Kathy Hochul signed S.9051B/A.10379C, which bars companies from offering "unsafe" AI chatbot features to users under 18 — such as pretending to be human, claiming a relationship with a child, encouraging self-harm, or urging kids to keep secrets. According to a Holland & Knight legal analysis, it takes effect January 1, 2027. Even outside New York, it is a handy checklist for any family reviewing the chatbots their kids use.
eSafety Commissioner's survey of 1,950 Australian children aged 10–17 (2 Feb–4 Mar 2026) found 78% had used an AI assistant and 8% an AI companion; ChatGPT was the most common. Among users of assistants or companions, 20% used them daily or more and 28% several times a week. Twenty percent experienced potentially inappropriate or harmful interactions, and 32% shared personal or sensitive data. Commissioner Julie Inman Grant: AI does not replace a trusted adult.
On November 5, 2026, Seattle Public Schools announced tighter digital guardrails: YouTube blocked for K–5, Minecraft blocked K–12 (except when part of the curriculum), and Gemini banned districtwide after closing a loophole students had used. The plan pairs less screen time in elementary grades with grade-by-grade AI literacy — no classroom AI in K–2; learning about AI from grade 3; guided bias and error work in grade 6; argument stress-tests and draft feedback in high school.
On iOS, iPadOS, macOS, watchOS, or visionOS 27 or later, Family Sharing lets a parent set Content & Privacy Restrictions → Siri to Siri AI (Beta), Siri Classic, or Don't Allow Siri. Siri AI (Beta) is only for children 13 and up; under-13s can use Siri Classic. Parents can also block writing help, math assistance, realistic images, and the ChatGPT extension.
On October 7, 2026, Common Sense Media's Youth AI Safety Institute rated ChatGPT for Teens an "Unacceptable Risk" after testing more than 4,000 prompts on accounts for ages 13–17. It found parent alerts, crisis referrals, and study mode fall short of what OpenAI promised, and called on OpenAI to restrict ChatGPT to 18+ until those protections are fixed. OpenAI's minimum age remains 13 (with a parent or guardian's permission under 18).
On November 3, 2026, Google announced that Guided Learning — an AI study mode that coaches students step by step with questions instead of giving final answers — is now available globally in Google Classroom and is coming to Google Search through a pilot in select U.S. schools. Google also said Gemini, Gemini Notebook, Gemini in Classroom and Search will be off for K-12 student accounts until school admins turn them on.
As of an October 29, 2026 update, Meta is rolling out alerts globally to parents who supervise their teens through Family Center: they're notified if their teen repeatedly searches suicide or self-harm terms on Facebook or Instagram, or if a conversation with Meta AI on Instagram, Facebook or Messenger suggests risk. Alerts don't reveal the chat itself and only work when supervision is turned on. Facebook's minimum age is 13.
In October 2026 OpenAI expanded ChatGPT's teen protections: parents can link accounts, schedule study mode, set quiet hours when ChatGPT is off, and get alerts about serious warning signs. Users must be at least 13, and under-18s need a parent or guardian's permission.
According to UNICEF (June 2026), at least 20 million children aged 12–17 across 10 countries have used AI, and they are more than three times as likely to use it as their parents. Homework and finding information are the top uses, and 1 in 10 turn to AI for advice about things that worry them.