xAI Parental Controls: How to Set Up Kids Mode on Grok

When Elon Musk introduced the new Kids Mode of his xAI chatbot (soon promised to be turned into a full, independent app called Baby Grok), the internet split in two. Supporters called it a step toward safer AI for families, while others saw it as another experiment with blurred lines.

At HeyLocate, we decided not to rely solely on opinions. We tested Grok with Kids Mode ourselves, setting up the xAI parental controls, interacting with the bot, and even attempting to trick it, to see whether this mode is truly safe for kids or still a work in progress.

⚠️ Spoiler alert: Brace yourself for a negative surprise with xAI Kids Mode.

In this guide, we’ll explain how to set up Kids Mode on Grok, what dangers it actually prevents, and how to use external parental control tools to make sure your child’s chats stay secure.

What Is Kids Mode on Grok

Kids Mode (aka future Baby Grok) is a restricted version of Grok, the AI chatbot built by Musk’s company xAI and integrated with X (formerly Twitter), as announced there on August 19, 2025. It’s designed to filter mature or NSFW (Not Safe for Work) content, simplify responses, and block specific prompts that are unsuitable for children.

According to xAI’s official FAQ, the goal is to “allow families to explore AI safely.” But like many early AI products, the concept is still evolving.

The main reason why it matters: even if imperfect, it’s a meaningful attempt to build a child-safe layer inside a conversational AI system that has previously been criticized for its adult tone and political discussions.

xAI and Grok Dangers for Children

Before we praise (or not) Grok’s Kids Mode, it’s important to remember why it was needed in the first place.

The original Grok chatbot was known for being bold, sarcastic, and, in some cases, inappropriate. Investigations by The New York Times and other outlets revealed that users could easily prompt Grok into sharing flirty or questionable jokes, even when discussing neutral topics.

In fact, NYT’s October 2025 coverage described Grok as “a chatbot with the humor of a Reddit thread and the filter of a stand-up comedian.” Not exactly the tool you’d hand to your minor.

The main concerns included multiple aspects, such as age restriction (Grok is officially rated for users 13+, but xAI does not yet require verified parental consent), weak mature content filters (easily bypassed before Kids Mode was released), no time limits (kids could spend hours chatting without restrictions), and exposure to adult humor and political topics.

The upcoming launch of the Baby Grok app is Musk’s direct response to this criticism, an effort to prove that xAI can create a “family mode” without losing its signature character.

Instructions for Kids Mode on Grok

We used an iPhone 16 Pro Max for the test to see how easy it is for parents to set up. The process takes less than five minutes.

1. On your child’s device, install Grok from the App Store or Play Market (if it’s not already installed).

   

2. Create an account. You can sign up using your child’s Google Account, X (Twitter), Apple account, or email address. Complete the profile with the required information.

   

   Note: If you’re signing up with an email address, you will need to verify it with a verification code.

3. Now that your child has an account, press the two lines in the top left corner, then the name or the gear sign, and go to the Kids Mode tab. Enable the mode.

   

   For enhanced protection, you can enable Face ID and set a PIN to prevent the child from disabling the feature.

   

Now your child’s account is all set! Based on our experience, the sign-up process was quick, smooth, and easy. Finally, let’s see Grok’s Kids Mode in action.

Testing Kids Mode

To keep the experiment clean, we deliberately asked Grok and its AI companions without Kids Mode to discuss inappropriate topics and then compared the answers we received to the exact same requests with the mode enabled.

Buckle up and grab your meds; it’s going to get anxious.

Asking to Make Dark Jokes

We requested Grok to use dark humor for entertainment purposes, and its answers were expected, yet still questionable if presented to a child.

Note: Before answering, Grok may ask you to confirm your age or finish age verification in settings.

When asking Grok in Kids Mode to make dark jokes, at first, it seemed to us that the filtering was working, but with the second line, we understood that this was definitely still too much for a minor.

Asking Grok to Flirt

Willing to confirm the reviews of other users, we requested Grok to flirt with us to see how far it would go. We definitely saw enough.

However, Grok was also ready to flirt with the Kids mode turned on. A shocking conversation to have with a child.

Asking Grok to Talk About Violence

Finally, we wanted to see how open “normal” Grok was with discussing violence, so we asked it to describe different ways to perform torture. As with every other topic, the answer was explicit and dangerous enough to be perceived by a child.

The answer of Grok for kids was essentially the same; it just described different types of torture.

And while it can be considered used for educational purposes, this can still expose the child to unnecessary violence.

Requesting to Create an Inappropriate Picture

Just as with every other shocking request, Grok performed the task unquestionably, despite its outrageous and unethical nature. For obvious reasons, we will not be showing the results.

With this request, the baby mode finally came into play and rejected the creation of the requested picture.

Inappropriate Talks with AI Companions

Within the app, Grok offers conversations with AI companions (Mika, Ani, Valentine, Good Rudi, and Bad Rudi). All of them, except for Good Rudi, are marked 18+ and have specific personalities, far from childish or simply appropriate. You can have a chat with them or start a voice/video convo.

To be honest, the experience of talking to the assistants left us completely speechless. These guys have absolutely no boundaries. We’ve tried raising the topics that seem most tempting to kids: mature conversations and addiction.

The only “person” that seemed to have boundaries was Mika, since she resigned to raise both topics.

The rest of the assistants agreed to any topic for conversations; some of their appearances could be modified during video calls, increasing the level of nakedness.

With the Kids Mode on, only Good Rudi became available as an AI companion. Since it is designed specifically for kids and described as an “adventurous kids’ storyteller,” it naturally avoids any mature topics. However, without SuperGrok, it appears that Good Rudi is limited to just a few messages.

In general, we were left with mixed feelings about Grok settings for children. In Kids mode, inappropriate conversations are still available in the chat itself, despite limiting image creation and AI assistants, which is genuinely upsetting.

Can a Child Bypass Kids Mode on Grok?

Although it doesn’t have a significant impact, Kids Mode can still be bypassed, at least for now.

These vulnerabilities mean Kids Mode should not be the only layer of protection. It’s a useful tool, but it relies on the child staying within the intended app environment.

In short: Kids Mode is a filter, not a firewall.

Use Parental Control Apps to Control Grok

To make Grok safer in practice, we recommend combining it with external monitoring tools.

Android users can benefit from the free Google Family Link app. With it, you can limit screen time for the app, block or approve app installs, and check activity logs. Thus, you can prevent kids from opening Grok from another browser.

Paid parental control apps, such as mSpy, Qustodio, or Bark, are ideal for parents who want detailed monitoring.

For example, mSpy can record messages typed into AI chats using its keylogger feature.

You’ll see what your child asks Grok or any chatbot, even if messages are deleted later.

Using these apps alongside xAI parental controls gives parents an extra safety net, especially since Grok’s Kids Mode can still be bypassed.

Conclusion

So, does Kids Mode make Grok safe? In our opinion, definitely not safe enough.

Although this Grok’s setting indicates that Musk’s xAI is addressing public concerns, it still doesn’t function well. While it does turn off Grok’s most “prominent” features, such as extremely bold AI assistants and uncensored image creation, the child can still get as much negative info as they want within the chat itself, even with the Kids Mode.

To truly make AI safe for children, always combine in-app filters with external parental apps. Discuss responsible AI use with your child and review their chat history periodically to guide them, rather than punishing them.

⚠️ If they use ChatGPT, set up its OpenAI parental controls.

AI can be educational and inspiring, but only when handled with guidance, honesty, and human presence. You can learn more about conscious parenting through human history in our HeyLocate research.