Skip to content
trending

Guides

Parental Controls in ChatGPT, Gemini, Copilot, and Character.AI

Learn how to set up parental controls in ChatGPT, Microsoft Copilot, Gemini, and Character.AI, manage quiet hours, and restrict AI features.

By 7 min read

👁 116 reads

You can enable parental controls in ChatGPT directly within Settings under the Parental controls section: send an invite to your child, and once they accept, you can schedule quiet hours, manage chat memory, voice mode, image generation, and study hours. With Gemini, Copilot, Character.AI, and Meta AI, controls work differently - some rely on Google Family Link or Microsoft Family Safety, while certain services enforce regional age restrictions.

This is part of our AI in education series, current as of October 12, 2026. If you are exploring this topic for the first time, check out our guide on what parents need to know when kids use ChatGPT for baseline age limits and communication tips. Below is the technical setup.

Comparison: What Parents Can Control

Service Age Limit Where to Configure Parental Controls Available Can Parents View Chats?
ChatGPT 13+, under 18 with parental consent In-app, account linking Quiet hours, memory, voice, image generation, model training opt-out, study hours, safety alerts No
Gemini Independent Google Account age varies (13 in US/UK, 14-16 in parts of EEA) Google Family Link Gemini Apps are completely unavailable for supervised accounts in the EEA, Switzerland, and the UK Not applicable
Copilot 13+ (higher in select regions) Microsoft Family Safety Block apps and websites, screen time limits, toggle image generation, voice mode, group chats No chat viewing feature documented by Microsoft
Character.AI Under 18 barred from open chats In-app (age assurance) Weekly parental insight email showing screen time and top characters if enabled by the teen No
Meta AI Teen accounts Instagram, Facebook, Messenger, WhatsApp AI character interactions paused for minors; parental supervision tools announced in select countries Not confirmed globally
Claude 18+ No dedicated teen tier Not applicable Not applicable

ChatGPT: Step-by-Step Parental Controls Setup

This feature has been active since September 29, 2025, and is available on Free tiers as well as paid plans. You will need two separate accounts: your own and your child’s.

  1. Sign in to your ChatGPT account (preferably in a desktop browser) and open Settings from your profile menu.
  2. Navigate to “Parental controls” and choose the option to add a family member.
  3. Enter your child’s email address or phone number and send the invitation.
  4. The child must accept the invite via the message received or inside their account settings. A teen can also initiate the invite to the parent.
  5. Select your child’s name in your family members list to configure the controls below.

Key toggles to manage:

  • Quiet hours - Set designated recurring time windows during which ChatGPT access is locked (for instance, 10:00 PM to 7:00 AM or school hours).
  • Study hours - During these set intervals, new chat sessions automatically initiate in study mode, guiding learning via Socratic inquiry instead of directly providing completed solutions. They do not lock standard chat access entirely and do not apply to custom GPTs or Projects.
  • Memory - Turning this off stops ChatGPT from personalizing responses using saved biographical facts about your child.
  • Voice mode and image generation - Both features can be disabled individually.
  • Model training - Opt out so your child’s chat inputs are excluded from OpenAI model training pipelines.
  • Safety notifications - In rare high-risk safety situations, OpenAI can trigger an alert to the parent. As of July 2026, this includes automated alerts if an account is suspended for violent threats.

What the feature does not do: parents cannot read chat logs, and a teen can link their account to only one parent. If the teen unlinks the accounts, the parent receives an alert. In addition, these settings do not apply if ChatGPT is accessed logged-out in a web browser - preventing that requires device-level controls.

Device-Level Restrictions: Phone and Desktop

Because native in-app settings only apply when your child stays signed into their linked account, your operating system serves as the second line of defense.

  1. Android: Use Google Family Link to enforce daily limits on chatbot apps and require parental approval before new apps are installed.
  2. iPhone and iPad: Configure Screen Time to set application time limits and add Content & Privacy Restrictions to block unauthenticated web access via mobile browsers.
  3. Windows and Xbox: Add your child’s account to Microsoft Family Safety, which controls web access and software permissions across the desktop environment.

While Google provides management options for Gemini via Family Link in select territories, official Google support documentation states that Gemini Apps are not available on supervised Google Accounts in the EEA, Switzerland, or the UK (Google Supervised Accounts documentation, verified October 12, 2026). In countries where the minimum age to manage an independent Google account is 14 to 16 (such as 16 in Germany and the Netherlands, 15 in France, or 14 in Spain, Italy, and Austria, per Google Account Age Requirements, checked October 12, 2026), minors below those age thresholds cannot access Gemini on personal supervised accounts.

The exception is school-managed access: Gemini for Education is included within Google Workspace for Education. However, it must be deployed and enabled directly by school IT administrators, and student data submitted through enterprise education accounts is excluded from training AI models.

Copilot: Microsoft Family Safety

  1. Add your child to your family group within Microsoft Family Safety.
  2. Block the Copilot application or web URL entirely, or set restrictive screen time allowances.
  3. If granting access, turn off specific features you wish to restrict: image creation, voice interactions, group sessions, and AI companion features.

According to Microsoft, accounts registered to users under 18 do not have their data utilized for AI model training, personal profile tailoring is restricted, and answers touching sensitive topics are filtered for age appropriateness. In academic environments, Microsoft 365 Copilot Chat is open to users aged 13 and up, but it remains disabled by default for primary and secondary institutions pending administrative review.

Character.AI and Meta AI

As of November 25, 2025, Character.AI restricts users under 18 from unrestricted open-ended character interactions. The platform estimates age using automated detection models, escalating flagged profiles to identity verification vendor Persona for document or video selfie checks. Minors are restricted to Stories and structured reading modes. Since March 2025, Character.AI has offered “Parental Insights” - a weekly summary email detailing overall screen time and frequently contacted character categories, without sharing chat transcripts. Note that this summary must be initiated by the teenager from their settings menu. For deeper background, see our breakdown on Character.AI features and age guidelines.

On January 23, 2026, Meta suspended open teen access to third-party AI characters across Instagram, Facebook, Messenger, and WhatsApp globally. The core Meta AI assistant remains accessible with baseline minor protections. Parental management tools specifically for Meta AI features were announced for the United States, United Kingdom, Canada, and Australia. For messaging settings, review our guide to WhatsApp parental controls.

Sample Prompt to Review Account Rules

Once you finish configuring account permissions, sit down with your teen and have them submit this prompt in their ChatGPT window. The resulting summary will clearly show both of you what restrictions are active. A teenager can query their account settings via the prompt, but cannot override the rules set by the parent.

Note: The following prompt is an English translation of the prompt tested in our original Polish benchmark:

I am [age] years old and my account is linked to my parent's account.
Explain to me in plain language:
- What rules apply to my account,
- What actions are blocked during quiet hours,
- How study mode works and how I can trigger it myself,
- Whether my parent can read my chat transcripts.
Answer in concise bullet points, in 10 sentences or fewer.

Frequently Asked Questions

Are ChatGPT parental controls free to use?

Yes. Parental controls are built into all tiers, including the Free version, for both the parent and teen accounts.

Can parents view their child’s ChatGPT chat history?

No. Neither ChatGPT nor Character.AI grants parents direct access to read message logs, and no such oversight tool exists for Copilot or Meta AI. ChatGPT only generates alerts for severe safety and violence-policy violations, while Character.AI provides high-level time summaries without chat text.

How do I block ChatGPT completely on a child’s device?

On Android, restrict it using Google Family Link; on an iPhone or iPad, enforce app and URL blocks through Apple Screen Time. Remember to block both the official app and chatgpt.com in mobile browsers, as logged-out users can still access web interfaces.

Can a teenager disable parental controls unilaterally?

Within ChatGPT, a teen can choose to unlink the accounts, but doing so triggers an immediate notification to the parent’s registered email. A child cannot modify individual restrictions set by the parent while linked.

Can minors access Google Gemini?

In the US and UK, users can operate an independent account starting at age 13. However, across the EEA, Switzerland, and the UK, Google Gemini is strictly unavailable on Family Link supervised accounts. Underage students in those jurisdictions can only access the tool if their school provides an enterprise Google Workspace for Education account with Gemini enabled.

Google · Your sourcesQuantslant higher in your Google - one clickAdd to preferred sources →

Sources

See also

Kinetyra

// read next

More on this from QUANTSLANT

Piotr Olszewski

Editor and publisher · Quantslant

I edit and publish Quantslant. AI tools help with research, drafts and translation; choosing topics, running the tests, checking facts and the decision to publish are mine, and so is the responsibility for what appears here.

read nextbenchmarksMuse Wonder vs Claude Opus 5.5, Fable 5.1 and GPT-6 Astra: BenchmarksMuse Wonder takes on Claude Opus 5.5, Fable 5.1, and GPT-6 Astra. Detailed benchmarks, independent index results, Polish-language testing, and API pricing.

Most read

  1. 1Best AI Video Generator Apps for Your Phone (2026)
  2. 2Claude Opus 5.5 Motion Graphics: 60s Showreel from One Prompt
  3. 3Wan 2.2 on RTX 4090: timings, VRAM and four test clips
// site map

This page in other languages6

Quantslant in numbers 1,427posts in archive26news from the last 7 days18video models in the observatory
🎬 AI Video Benchmarks: Higgsfield DoP vs OpenAI Sora • Higgsfield vs Runway Gen-3 • AI Video Credit Calculator
× ‹ enlarge ›