OpenAI has launched a special ChatGPT mode for teenagers

OpenAI has launched a special teen mode for ChatGPT with additional content restrictions, parental control features, and a user age detection system. The mode is designed for underage ChatGPT users, primarily teenagers aged 13–17.

The company emphasizes that this is not a separate app, but a special mode of ChatGPT. If the system determines that the account owner is under 18, it may automatically enable additional protective measures and some feature restrictions.

ChatGPT is not intended for children under 13. Users aged 13 to 18 need parental or guardian consent to use the service.

ChatGPT will try to determine the user's age

One of the key elements of the system is age prediction. OpenAI may analyze signals associated with the account and determine whether the user is likely underage.

Among such signals, the company names common topics of conversation with ChatGPT, time of day when a person uses the service, account usage patterns, and account age.

The system may apply teen mode even if the user has previously self-reported a date of birth.

If the system incorrectly identifies an adult user as a teenager, the person can verify their age and disable the additional restrictions. For age verification, OpenAI uses the Persona service. Depending on the country, the user may need to provide a selfie or an identity document.

OpenAI states that it does not receive the uploaded document or photo. The company only receives the necessary age information. Persona must delete the verification materials within seven days.

The age prediction system is being rolled out globally. In European Union countries, its deployment will occur separately, considering regional requirements.

What ChatGPT responses will be restricted for teens

In teen mode, additional sensitive content filtering is enabled by default. ChatGPT will respond more cautiously to requests involving naturalistic violence, dangerous internet challenges, and other potentially harmful content.

Additional restrictions also apply to sexual, romantic, and violent roleplay scenarios, as well as content that promotes extreme appearance standards, unhealthy weight loss, or humiliation of a person because of their body.

Teens will still be able to use ChatGPT for studying, information seeking, creativity, and other normal tasks. OpenAI also states that accounts in teen mode will not be shown advertisements.

Parents will be able to set "quiet hours"

Another block of changes concerns parental control. A parent or guardian can link their account to the teen's account and manage some of its settings.

In particular, they can enable or disable voice mode and image generation, manage the use of saved memory, data-sharing settings for model improvement, and additional filtering of sensitive content.

Parents can also set so-called quiet hours — a specified time period during which the teen cannot use ChatGPT. For example, this can restrict service use at night or during study time.

Additionally, through parental control, they can activate Study Mode. In that case, new conversations of the teen with ChatGPT will start in study mode by default.

Parents cannot read teens' conversations

OpenAI separately emphasizes that connecting parental control does not give parents access to the teen's conversation history and does not allow monitoring their correspondence in real time.

Exceptions are specific situations when safety systems and OpenAI specialists identify signs of serious risk to the teen. In such cases, the linked parent or guardian may receive a notification, but the company promises to share only the information necessary to ensure the child's safety.

The teen can also unlink the account from parental control on their own. In that case, the parent will receive a notification.

When the system determines that the account owner has turned 18, teen restrictions are lifted, and the connected parental control ceases to operate.

Sources: OpenAI — Age prediction in ChatGPT, OpenAI — Parental Controls