Affiliate Disclaimer ChatGPT and Sora settings now have parental controls that any parent can set up for a linked child’s account. However, that was a forced measure made by OpenAI after the tragic case of Adam Raine, a young teen who ended his life after engaging in troubling conversations with ChatGPT in April 2025. This, in itself, is proof of why AI parental controls are necessary.
As a parent, it’s natural to worry about what your child might encounter online, but with the rise of AI, these concerns now extend beyond traditional platforms, such as social media and YouTube. The truth is, even when I blocked my daughter’s phone access to browsers with Family Link, there’s no option to block AI chats available, for example, through WhatsApp.
That’s why OpenAI parental controls are what we need now to guide our children’s inevitable AI interactions for a safer, more age-appropriate experience. I tested how these settings work and prepared this guide on how to best manage your child’s use of ChatGPT and Sora.
OpenAI Age Restrictions
ChatGPT is not intended for children under the age of 13. Moreover, OpenAI has established strict guidelines requiring children aged 13 to 18 to obtain parental approval before using the platform.
Indeed, when creating an account for a child, ChatGPT did not allow it when the age was under 13. However, I simply changed the year of birth, and it worked. Thus, the system does not verify the age of its users in any way, but still requests parents’ consent if your Google accounts are linked.
Additionally, OpenAI is developing an age prediction system that should help automatically determine whether a user is under 18 based on their interactions with the platform. In this case, ChatGPT will automatically apply age-appropriate settings. When the system cannot confidently predict a user’s age, OpenAI plans to request identification documents to verify the user’s age.
The system is still under development, and everyone is waiting to find out how it will work. It is even intriguing to see how many adults may be mistaken for minors by AI.
What Are Real AI Dangers for Children?
The dangers of AI for children go far beyond the usual risks of inappropriate content like graphic violence or explicit material. In fact, if you use ChatGPT, then you know that it is quite conservative in this sense and doesn’t show anything rated 18+. However, not every AI chat is like that.
The potential hazards associated with AI chats are numerous. A study analyzing over 150,000 user reviews identified approximately 800 instances where users reported unsolicited sexual content and predatory behavior from the chatbot.
Not to mention the numerous lawsuits filed by parents accusing chatbots of inappropriate communication with their children, encouraging disobedience, and even suicide, as in the case mentioned at the beginning.
It’s already a fact that AI agents can talk sexually or suggestively to children, often through roleplay or direct conversation. These conversations could be inadvertently triggered when a child asks for help with “adult” topics or uses specific keywords that prompt inappropriate replies from the AI. Parents know how children learn to communicate with high-tech devices better than anyone (obviously, even better than the carefree AI developers).
Meanwhile, research proves that AI is becoming an increasingly common tool for kids as young as 8 or 9 years old, with many using it for homework help or casual conversation. A survey by Kaspersky in 2025 found that children aged 8–10 spend an average of six hours daily on screens, with a notable increase in interest towards AI tools.
Despite the growing use, many parents remain unaware of their children’s interactions with AI. A study by Common Sense Media indicated that 7 out of 10 teens use generative AI tools, yet only 37% of their parents are aware of this usage.
However, if you are here, you are already conscious of the necessary steps to be done to control AI (ok, at least to try to) instead of allowing it to control your children. Thus, let’s go straight to precise instructions.
How to Set Up ChatGPT Parental Controls
To put parental controls on the ChatGPT your child uses, first, you need to link your OpenAI accounts. Here’s a step-by-step guide to my experience and how you can do this as well.
- Log in to your OpenAI account on the ChatGPT website.
- Navigate to Settings: tap the burger menu in the upper-left corner of the phone screen, then your name, and choose Settings.
On PC, navigate to your name in the bottom left corner. - You’ll see the Parental Controls feature and the option to Add a family member. Tap it to invite your child through their email address or phone number.
- Your teen needs to accept the invite for you to be able to adjust their settings and time limits. They’ll see the notification in their ChatGPT. OpenAI states they won’t share details of kid’s conversations, except in certain rare safety circumstances.
- After the child accepts the invite for linking your accounts, you’ll see them in the Parental Controls section of your account. You can also add other children there.
- When a parent’s account is linked with their teenager’s, the teen’s profile is automatically placed under extra safeguards. These include reducing sensitive content with stronger filters that limit exposure to graphic material, risky viral challenges, sexual or romantic roleplay, violent themes, and unrealistic beauty standards.
Other settings focus on the general user experience and privacy options, such as allowing OpenAI to use a child’s transcripts and files from chats and enabling minors to use voice mode and image generation.
When I turned off the image generation option, ChatGPT suggested my teen design a prompt to use in another genAI tool instead. In other words, it’s not a very reliable parental control feature.
The only other option for real parental controls is to set quiet hours when your child can’t use ChatGPT.
Thus, your teen will receive a notification about the time for a break in the set hours.
Please note that your youngster will receive notifications about any parental control changes you make.
Turn On Teen Safety Notifications
Parents cannot view their teens’ messages with ChatGPT, but, after suicides with the help of AI, OpenAI decided to use their system and trained reviewers to identify potential serious safety concerns in teens’ conversations. In these cases, parents may be alerted.
Ensure Teen Safety Notifications are on in your account’s Settings > Notifications. Decide whether you want to receive alerts via email and/or push.
With these alerts, OpenAI promises to provide only essential information about teens’ safety and resources to help parents support their children. Thus, you may receive:
- a notification if a possible self-harm risk is detected;
- links to mental health resources, crisis helplines, and guidance.
Please note that these notifications are not a substitute for professional medical care or emergency services. They are meant to help families stay informed and provide timely support when needed.
How to Manage Sora Parental Controls
In addition to ChatGPT, OpenAI also offers some parental controls for its Sora, a new AI video creation app. Since Sora involves multimedia content, it has its own set of concerns that differ from those of ChatGPT. However, the parental control tools offered are far from comprehensive.
- Personalized Feed: By default, Sora curates a personalized feed based on a user’s previous interactions. You can toggle this feature off to prevent your child from receiving content based on their activity.
- Continuous Feed: This feature allows an uninterrupted scroll of videos. You can disable this to prevent endless scrolling and ensure that your child engages with content mindfully.
- Messaging: You can disable messaging in Sora, ensuring that your teen cannot send or receive direct messages from other users.
Sora, like TikTok or YouTube, can expose children to viral trends or even harmful challenges, which is why it is essential to manage at least these settings for a safer kids’ experience there.
Can a Child Bypass OpenAI Parental Controls?
Unfortunately, yes, it’s easy for a child to bypass OpenAI parental controls if they simply unlink their account from their parents. The good news: you will be notified if this happens.
Otherwise, teens may find ways to access ChatGPT or Sora from other devices or accounts where parental controls are not set up.
Moreover, a savvy teen may learn how to ask the AI for content without triggering the parental filters directly. Thus, we need some time to check how effective these controls will work.
The main question remains open: if AI is so smart that it can replace certain professions, why can’t it become a real digital guardian for our children?
What Should Parents Do Instead?
- Keep communication open: Discuss the potential risks with your teen and clarify what kinds of conversations are inappropriate.
- Set family rules: Establish screen time limits, guidelines for online behavior, and ensure your child understands the importance of safe online habits.
- Use additional parental control tools: Consider using network-level or device-level parental controls to supplement the OpenAI settings. Check out our top family trackers. Some of them, such as uMobix, allow you not only to set restrictions but also to view screenshots of your child’s phone browser usage and see keylogger captures.
Conclusion
OpenAI implemented parental controls, but it seems like it’s only to calm the storm over ChatGPT’s impact on teens’ suicides. Like, they did something. The only thing that can be said in their defense is that no digital parental control will help if there is no digital education, trust, and open communication with the child at home.
Still, setting parental controls for ChatGPT and Sora is a minimum that parents should do. Now you know how.
⚠️ Moreover, remember that these AI tools are only two out of a big and growing world of artificial intelligence. Check which one your child uses to discover how you can control that, then.
With the right combination of tools and discussions, you can more confidently guide your child through this new world of AI.
FAQ
To create and use a ChatGPT account, you generally need to be 13+ years old (or 16+ in the EU/UK), and if you’re under your country’s age of majority (often 18), you need a parent’s or guardian’s consent.
While ChatGPT is safe overall, there have been reports and even lawsuits where it gave harmful responses to vulnerable teens, including advice or notes about self-harm. Studies also found it can sometimes provide dangerous content if safety filters are bypassed. Teens may become emotionally reliant on it, which increases risk.
Parental controls on AI phones depend on the specific device and platform. Look for parental control options within the device’s settings or use third-party apps for additional security.
Yes, there are AI tools made specifically for kids, with extra safety and learning features. Examples include Kai by Kinzoo (creative tools inside a safe messaging app), Hachiko AI (child-safe chat powered by GPT-4), GPTeens (for teenagers, tied to school learning), and research projects like Scratch Copilot (helps kids code) or Capybara (AI + AR creative play).
However, even “kid-friendly” AIs can slip up. Risks include misinformation, inappropriate answers, or kids forming over-reliance on the AI.
The safest options give parents visibility and controls. These tools can be fun and educational, but they should be used with guidance and supervision, not as a replacement for real teachers or caregivers.
Currently, there isn’t a specific “kids’ version” of ChatGPT, but the parental controls provide a way to tailor the experience to be safer for teens.
Monday-Sunday: 5:00am - 8:00pm (PT)
Leave a Comment