OpenAI released ChatGPT for Teens in August, and Engadget has published a detailed guide to the parental-control layer that sits around it. According to Engadget, OpenAI released the teen-focused experience after almost a year of development, positioning it as an education-oriented version of ChatGPT with tighter content controls and features meant to encourage healthier use by young people. The default experience does not hand parents full control automatically. Engadget reports that OpenAI says it automatically places users it believes are under 18 into ChatGPT for Teens, but the guide notes that several of the more consequential settings require a parent or guardian to link their own ChatGPT account to the teen’s account. Engadget also says child safety experts believe the default restrictions on many technology platforms do not go far enough on their own. The account-linking flow runs through ChatGPT’s settings, according to Engadget. A parent opens the Parental controls menu, chooses whether to invite the teen by email or phone, enters the teen’s contact information and marks the invited person as “My child.” After the teen accepts the invitation, the child’s name appears under Family members, where the parent can review and adjust that account’s settings. Engadget says the flow can also run in reverse: a teen can invite a parent or guardian by choosing that relationship instead. The controls are age-limited. Engadget reports that parental controls remain available only while the child is under 18. If OpenAI determines that a ChatGPT for Teens user has become an adult, the company notifies both the parent and the child, disables any parental controls and disconnects the accounts. After that, the parent can no longer manage settings or receive teen safety notifications tied to that account. Engadget identifies several of the settings parents will see once accounts are linked. “Reduce sensitive content” applies stronger safeguards intended to limit sensitive or age-inappropriate material, and Engadget says it is enabled by default. “Improve the model for everyone,” which controls whether OpenAI can use the teen’s conversations to train future ChatGPT models, is also enabled by default, according to the guide. “Reference saved memories,” which lets ChatGPT personalize answers based on prior conversations, is enabled by default in some regions depending on local regulations. The practical takeaway is that OpenAI’s teen mode is not a single switch. It combines automated under-18 enrollment, default restrictions and a separate parent-linked management layer. For families setting up ChatGPT for a teen, the linked-account setup affects how much oversight a parent can exercise. Who benefits: Parents and guardians benefit from a clearer path to manage a teen’s ChatGPT settings. Who's exposed: Teens whose parents do not link accounts may remain dependent on OpenAI’s default restrictions. Parents who assume teen mode alone provides full oversight may miss controls that require active setup.