AI News Feed
Market watch
Products & Applications

How to set up ChatGPT's parental controls to protect your teen

Step-by-step guide to enabling parental controls for ChatGPT for Teens, including account linking and key settings.

Child safety experts quoted by Engadget say the default restrictions do not go far enough. The ability to manage the more protective settings starts with linking the two accounts.

Account linking is done by going to Settings and selecting Parental controls, which appears under App settings on mobile. The parent selects Add family member, sends an invitation by email or phone, and then identifies the recipient as 'My child' when prompted. After the teen accepts, the child's name is listed under Family members, where the parent can review or change individual options. Teens may also invite a parent by choosing 'My parent or guardian' in the same relationship field.

The link remains active only while the child is under 18. Should OpenAI decide that a user has become an adult, it moves the account to standard ChatGPT, informs both parties, disables all parental controls, and removes the connection between accounts. Afterwards, the parent can no longer manage settings or receive teen safety notifications.

Once accounts are linked, parents can see close to a dozen settings. 'Reduce sensitive content' is enabled by default and applies stricter safeguards against age-inappropriate material. 'Improve the model for everyone' is also turned on by default, permitting OpenAI to use the teen's conversations for model training; that toggle can be switched off. The 'Reference saved memories' option controls whether ChatGPT personalizes responses based on past conversations, and its default varies by local regulation.

Feature access is handled through additional switches. Codex network access, cloud browser, Voice Mode and image generation all default to on, allowing teens to use them unless disabled. The educational 'Study Mode' is off by default; when enabled, it starts a new conversation in a format that encourages the teen to solve problems independently. 'Study hours' and 'Quiet hours' are also off until parents define schedules, respectively setting windows for study-oriented chats and blocking all use during specified times.

OpenAI states that switching off a setting hides or blocks the related feature for the teen, while switching it on makes the feature available under other safety protections that remain in place. While accounts are linked, teens cannot change settings controlled by a parent. The Engadget guide is intended to help parents decide which features are most suitable for their child's first experiences with AI.