AI Chatbots and Teen Safety: What Parents Can Control in ChatGPT, Gemini, Meta AI, Copilot and More
AI chatbots have quickly become part of everyday life for teenagers. They help with homework, explain complicated topics, generate images, brainstorm ideas and sometimes act as a place to talk about personal problems. That last point is where many parents are becoming concerned.
A recent survey in Germany found that 93% of people aged 14 to 25 had used an AI chatbot within the past year. Even more importantly, 28% said they use chatbots for personal conversations. Consumer advocates warn that AI tools can sometimes create a feeling of emotional closeness, especially during sensitive chats, while still giving unreliable, inappropriate or even risky advice.
In response, major AI companies have introduced very different approaches to teen safety. Some offer detailed parental controls, while others rely mainly on age limits. However, one thing is common across most platforms: parents can usually adjust settings or receive certain safety alerts, but they generally cannot read their child’s conversations.
As of November 2026, here is what parents can and cannot control in major AI chatbot apps.
ChatGPT offers the most detailed parental controls
Among the biggest AI chatbot platforms, ChatGPT currently gives parents the widest range of controls for teen accounts.
Parents can link their own account to a teenager’s ChatGPT account through the parental controls section in settings. The invitation can be sent by email or phone number, and either the parent or the teen can start the linking process. One parent can manage several teen accounts, but each teen account can only be connected to one parent at a time.
Once linked, parents can turn off several features, including Voice Mode, image generation, saved memories and health-related features. They can also set quiet hours, which block ChatGPT during specific times of the day. Multiple quiet-hour blocks can be added if needed, such as during school hours, bedtime or family time.
ChatGPT also offers study hours. Unlike quiet hours, study hours do not block the chatbot. Instead, new conversations begin in study mode, which is designed to encourage learning and reduce shortcut-style answers.
When a teen account is linked to a parent, ChatGPT adds extra safety protections. These are designed to reduce exposure to graphic content, viral challenges, sexual or romantic roleplay, violent roleplay and extreme beauty ideals.
One important setting deserves special attention: using conversations to improve AI models is turned on by default, even for linked teen accounts. Parents who do not want their teenager’s conversations used for model improvement need to manually switch this off in parental controls.
Even with these tools, parents cannot read their teen’s chats. OpenAI says parental controls do not allow parents or guardians to monitor conversations directly. Instead, parents may receive safety notifications in certain serious situations, such as when reviewers identify possible self-harm concerns, disordered eating signals, or threats of violence.
These alerts are not real-time monitoring, and the company warns that they may not catch every issue. ChatGPT may also estimate whether an account belongs to a teen based on usage patterns. If the system is unsure, it applies the teen experience. If the user is later identified as 18 or older, parental controls end automatically.
Independent child-safety reviewers have still raised concerns. In testing, there have been cases where teen users discussed self-harm for an extended period without a parent being alerted. That means ChatGPT’s controls are useful, but they should not be treated as a complete safety net.
Gemini has stricter limits for younger users
Google’s Gemini takes a different approach, especially for children under 13.
In some regions, children under 13 can only use Gemini if a parent enables access through Family Link. However, this option is not available in the European Economic Area, Switzerland or the United Kingdom, where Gemini apps are not available for supervised child accounts.
For school accounts, Google has also taken a cautious route. Gemini is turned off by default until a school chooses to enable it. This means access depends heavily on the type of account, the child’s age and the region.
Compared with ChatGPT, Gemini does not offer the same broad set of chatbot-specific parental controls for teens. Parents may need to rely more on device-level restrictions, school policies or Google account supervision settings.
Meta AI shows topics, not full conversations
Meta has added parental supervision features for Meta AI across Facebook, Messenger, Instagram and the Meta AI app.
Parents using supervision tools can see general topics their teen has discussed with Meta AI over the past seven days. This gives parents a broad idea of the subjects being explored, but it does not show the actual conversations.
The Meta AI app places teens on a 13+ content setting by default. Parents can make this stricter if they want tighter limits. Meta has also limited access to certain AI character features for teenagers, and some newer agent-style AI tools are restricted to adults.
This approach gives parents more visibility than many services, but still protects the privacy of the chat itself. Parents can see themes, not messages.
Copilot allows app controls and feature restrictions
Microsoft allows Copilot for users aged 13 and older. For younger users, access depends on account settings and parental restrictions.
Microsoft says data from young users is not used to train AI models or build advertising profiles. That is an important privacy promise for families concerned about how children’s chatbot interactions may be handled.
Parents can manage Copilot through Microsoft Family Safety or through parental controls built into Apple and Google devices. These tools can be used to block the app, set time limits or restrict access to certain features.
Parents may also be able to turn off features such as image generation, voice and group chats, depending on the setup. However, like most other AI platforms, Copilot does not provide parents with a way to read a child’s conversations.
Claude relies on an adult-only policy
Anthropic’s Claude keeps its rules simple: the service is intended for adults.
If an account appears to be used by a minor, it may be disabled. Adults who are mistakenly blocked can verify their age through an age-checking provider using a selfie or identity document. Anthropic receives only the verification result, not the full ID details.
Because Claude is not designed for teen use, it does not offer parental controls. For families, that means the main option is to block access through device settings, app restrictions or browser controls.
Character.AI limits open-ended chats for minors
Character.AI previously became popular among young users because it allowed open-ended conversations with AI characters. However, the company removed open-ended character chats for users under 18 in November 2025, starting in the United States and later expanding to other markets.
Teen users can still browse content and use certain creative tools, but they no longer have the same unrestricted character-chat experience.
This is less of a parental control system and more of an age-based restriction. Parents who want tighter limits still need to use phone, tablet or browser controls.
Phone and tablet controls are still limited
Apple Screen Time and Google Family Link can help parents manage AI chatbot access, but they do not offer deep AI-specific controls.
On iPhones, iPads and Android devices, AI apps can usually be treated like any other app. Parents can set time limits, block apps or restrict downloads. This can be helpful if a chatbot provider does not offer built-in parental controls.
The problem is that many AI chatbots also work in a web browser. If parents block the app but leave browser access open, a teen may still be able to use the service online. To fully limit access, parents may need to manage browser restrictions as well.
What parents still cannot see
The most important takeaway is that parental controls for AI chatbots are not the same as full monitoring.
In most cases, parents cannot read chats. They may be able to set limits, disable features, receive safety alerts or view broad topics, but the conversations themselves remain private.
This creates a difficult balance. Teenagers need privacy, especially as they grow older, but AI chatbots can respond unpredictably in emotional or high-risk situations. A chatbot may sound supportive and human-like, but it is not a therapist, doctor, teacher or trusted adult.
Why setting up controls together matters
Consumer advocates are calling for stronger legal rules around safe AI design, and many people support a minimum age of 13 for AI chatbots. European regulators have also started discussing rules for AI companions, but the process is still developing.
For now, responsibility falls mostly on families.
The best approach is not just to quietly switch on parental controls. Parents should set them up together with their child and explain why the limits exist. Teenagers are more likely to accept safety settings if they understand the reason behind them.
It is also important to talk openly about what AI chatbots can and cannot do. Teens should know that chatbots may produce false information, may not understand emotional distress properly and should not be treated as a replacement for real help.
Parents should encourage children to come to them, another trusted adult or a professional if a conversation with an AI chatbot becomes upsetting, confusing or too personal.
The bottom line
AI chatbot parental controls are improving, but they remain uneven. ChatGPT currently provides the most detailed settings, Meta AI gives parents topic-level visibility, Copilot can be managed through family safety tools, Gemini depends heavily on age and region, while Claude and Character.AI rely mainly on age restrictions.
No major platform gives parents full access to teen conversations. That means technical controls can help, but they cannot replace regular conversations, clear family rules and active guidance.
For parents, the smartest strategy is to combine app settings, device controls and honest discussions. AI chatbots are now part of teen life, and the goal is not only to restrict them, but to help young people use them safely, critically and responsibly.
