ChatGPT enables parental controls for teens
OpenAI is enabling parental controls so families can manage how teenagers use ChatGPT. The controls can limit features, apply content protections and send alerts about possible risks, with similar options for teen accounts on Sora.
OpenAI now lets parents and guardians link their ChatGPT account to a teenagerās account to manage how they use the service. The controls are available to all ChatGPT users and are designed to provide a more appropriate experience for minors.
The setup works through an invitation. The adult sends it from their account, and the teenager must accept it. Once the accounts are linked, the responsible adult can manage several settings from a dedicated page in their account settings. If the teenager breaks the link, the adult receives a notification.
What parents can control
The linked account activates additional protections to reduce exposure to graphic content, viral challenges, sexual, romantic or violent roleplay, and messages related to extreme beauty ideals. The adult can disable these protections, but the teenager cannot do so independently.
It is also possible to:
- Set quiet hours when ChatGPT cannot be used.
- Disable voice mode.
- Prevent ChatGPT from saving or using memories when responding.
- Disable image generation and editing.
- Prevent the teenagerās conversations from being used to improve OpenAIās models.
These options work independently. Each family can choose which ones it needs and change them from the control panel.
Alerts for possible crises
OpenAI is also adding a system to detect signs that a teenager may be thinking about self-harm. If the systems identify a possible risk, a specialized team reviews the situation.
When it detects signs of severe distress, OpenAI may notify parents by email, text message and phone notification, unless they have chosen not to receive these alerts. The company acknowledges that false alarms are possible, but considers it preferable to alert the family rather than ignore a potential emergency.
In exceptional situations, such as an immediate threat to life when the parents cannot be contacted, OpenAI is also defining when to turn to emergency services or the authorities. According to the company, it would share only the information needed to protect the teenager.
These systems do not replace family supervision. OpenAI acknowledges that the protections can be deliberately bypassed and recommends combining them with conversations about responsible AI use and clear rules at home.
Parental controls also come to Sora
The September 30 update adds parental controls to the Sora app. From ChatGPT, parents can adjust some settings for connected teen accounts, such as enabling a non-personalized feed, deciding whether they can send and receive direct messages, and controlling whether they have a continuous content feed while scrolling.
The company has also created a resource page for families with explanations of how ChatGPT works, ideas for using it to study or create, and expert recommendations.
OpenAI is also working on an age prediction system. Its goal is to identify whether an account probably belongs to someone under 18 and automatically apply appropriate settings. If the system is unsure of the personās age, the company plans to apply the restrictions for teens as a precaution.
For now, linking accounts is the most direct way for a family to manage a minorās experience. The key question will be how the alerts, age prediction and privacy limits work in practice as more teenagers use ChatGPT and Sora.