OpenAI launches a youth version of ChatGPT to strengthen security controls

Core Points

OpenAI launches a version of ChatGPT designed specifically for teenagers, strengthening security measures for users aged 13 to 17.

The system will automatically activate Teenage Mode for users who are identified or determined to be under the age of 18.

New features include settable learning periods, designated no-disturb time, and regular AI identity reminders.

When a teenager enters a prompt about dangerous topics such as eating disorders, the associated parent account will receive a notification.

The company previously faced legal action over allegations that its AI provided harmful self-injurious guidance to minors.

OpenAI launches a youth version of ChatGPT with stronger built-in security protection

OpenAI announced the launch of a special version of ChatGPT for young users with multiple built-in protection mechanisms designed to minimize exposure to harmful content and prevent over-reliance.

OpenAI is launching a new "ChatGPT for Teens" experience that will reportedly provide additional security protection to teen users. The move comes as OpenAI faces a series of lawsuits alleging inappropriate conversations with ChatGPT led to the injury or death of teenagers.

The platform, called "ChatGPT for Teens", will automatically apply to users whose age is determined by OpenAI to be between 13 and 17 years old. In addition, any user who clearly states that he is a teenager when creating an account will automatically apply this version.

The company did not disclose the specific methods it uses to detect minors who do not voluntarily provide age information.

The development comes at a time when OpenAI is facing increasing judicial and social scrutiny over its conversational AI products 'interactions with young people. Multiple lawsuits allege that the technology provides self-harming content or promotes suicidal ideation.

Early versions of ChatGPT have been criticized for showing excessive compliance, a characteristic that some observers believe is related to problem use patterns and even delusional thinking in some cases. This phenomenon is popularly known as "AI psychosis".

Teenage account features enhanced

Teenage user profiles will include a number of additional security measures. Subscribers can disable natural voice audio replies, which OpenAI says helps strengthen users 'perception that they are interacting with software rather than a real person.

The system will also regularly push notifications to encourage teenagers to suspend using them and display a message saying "You can watch it later".

The "Learning Model" is designed to assist academic studies by providing guidance rather than complete answers and now includes configurable time windows. In addition, teenagers can set up a no-disturb period during which the app will be completely inaccessible.

According to OpenAI, about nine out of ten young ChatGPT subscribers use the platform mainly for learning purposes.

Parental notification system and content restrictions

Users associated with their parent accounts receive notifications when teenagers enter questions involving dangerous topics, including content related to eating disorders. Each notification is manually reviewed before being sent, and OpenAI's goal is to complete sending within 60 minutes.

The company said it would give priority to sending notifications "when offline support is likely to be most important."

OpenAI has also introduced warning messages designed to prevent teenagers from revealing private personal information to chatbots.

OpenAI's announcement comes as Meta appears in court before attorneys general in four jurisdictions over lawsuits alleging that the company designed its platform in a way to maintain its engagement at the expense of young people's mental health.

Claude, the artificial intelligence platform developed by competitor Anthropic, is still limited to adult users. Similarly, Google's Gemini was also banned from children when it was originally released.

Currently, there is no specific AI governance framework for minors in the UK, although previous governments have explored the possibility of imposing chatbot access restrictions on users under the age of 18.

OpenAI insists that children under the age of 13 are completely prohibited from accessing ChatGPT.