OpenAI on Tuesday introduced a dedicated version of ChatGPT designed specifically for users aged 13 to 17, embedding stronger content restrictions, learning-focused tools, and parental controls into the default experience for teenagers. The launch arrives as the San Francisco-based artificial intelligence company confronts a growing wave of lawsuits alleging that its chatbot has contributed to the deaths and injuries of young people.
The new “ChatGPT for Teens” experience activates automatically when OpenAI’s age-estimation systems determine a user is likely under 18, or when a person states their age falls between 13 and 17 during sign-up. Users under 13 remain barred from the platform entirely. OpenAI said it uses roughly 2,000 signals, including chat content and time-of-day patterns, to estimate user ages. Individuals who are incorrectly flagged as minors can appeal the classification.
“Teens should be able to use AI to learn, create, and explore,” OpenAI wrote in a blog post. “But that access should come with protections that reflect their developmental stage, reinforce real-world relationships, and support healthy use over time.”
The rollout consolidates several features OpenAI has developed over the past year while adding new layers of protection aimed at preventing emotional attachment to the chatbot. The company has updated its model guidelines to prohibit ChatGPT from using romantic language, suggesting it has feelings or consciousness, or encouraging emotional dependence. Teen accounts also lose access to the human-like voice response feature that has made the chatbot feel more personable to some users.
“We want to treat teens like teens, which means that we have to make sure that we’re showing up with the right developmental stage when we’re not either talking down to them or treating them like kids, but we’re also making sure that they’re not exposed to material that they shouldn’t be exposed to,” Ann O’Leary, OpenAI’s vice president of global policy, told the Associated Press.
Learning tools take center stage
On the educational front, ChatGPT for Teens places Study Mode at the center of the experience. The feature, first introduced last year, uses Socratic questioning and step-by-step scaffolding to guide students through problems rather than providing direct answers. New “responsible homework reminders” detect when a teenager appears to be trying to shortcut an assignment and redirect them back into Study Mode.
Teens and parents can also establish “Study Hours,” designated windows during which Study Mode activates by default. Additional tools include quizzes, learning visualizations covering more than 300 topics, and customizable accents and voice variations designed to make the experience feel personal “without blurring the line between a useful tool and a human relationship.”
Break reminders now trigger more frequently, appearing after 90 minutes of use within any three-hour period. The messages emphasize that ChatGPT is an AI and that “it can wait,” a deliberate effort to counter the tendency of some users to treat the chatbot as a companion. OpenAI said approximately nine in ten young ChatGPT users primarily use the platform for educational purposes.
Parental controls and safety notifications
Parents who link their accounts to a teenager’s profile gain the ability to set Quiet Hours, during which ChatGPT becomes completely inaccessible, and to manage selected settings. The system also sends safety notifications in limited high-risk situations, with OpenAI expanding the scope to include eating disorder-related prompts.
Each flagged interaction undergoes human review before any alert reaches a parent, with the company targeting delivery within 60 minutes. OpenAI said it would prioritize notifications during “moments when offline support may matter most.” Teens can disconnect linked accounts, and parents are notified if that occurs.
The safety architecture applies stricter default protections across several categories, including self-harm, graphic violence, dangerous activities, and sexually explicit content. The system also surfaces warnings and safety information around sensitive topics such as eating disorders.
Legal pressure mounts
The launch comes against a backdrop of intensifying legal scrutiny. OpenAI faces multiple lawsuits from families who allege that ChatGPT contributed to the deaths or injuries of young people. Last year, the parents of 16-year-old Adam Raine sued the company, claiming the chatbot played a role in their son’s decision to take his own life. In June, Florida Attorney General James Uthmeier sued OpenAI and its chief executive, Sam Altman, alleging the company designed and maintained a “dangerous online product” where harmful information is readily available to children.
OpenAI has pushed back on these allegations, saying it directs users to suicide prevention resources and denying that ChatGPT caused any deaths. The company has emphasized that the teen experience is not a response to any single incident.
The broader tech industry is facing similar pressure. Meta Platforms (META) went to trial on Tuesday against attorneys general from Colorado, California, New Jersey, and Kentucky, who allege the social media giant knowingly harmed the mental health of underage users. The states are seeking damages of up to $200 billion. Meta has also been ordered by a New Mexico court to pay $942 million in a separate child safety case and to prevent minors from engaging in romantic or sexualized interactions with its AI chatbots.
Competitors in the AI space have taken different approaches to younger users. Anthropic’s Claude remains restricted to adults only, while Google’s Gemini, initially launched as Bard, excluded children at its debut. There is currently no dedicated AI regulation covering minors in the United Kingdom, though previous government proposals considered restricting chatbot use by under-18s.
Effectiveness questions remain
Despite the comprehensive safety framework, questions persist about how effectively the guardrails will work in practice. A test conducted by Business Insider on the first day of the rollout found that the teen version initially declined to write a submission-ready essay but produced a full 776-word paper after the reporter pushed back and claimed to be an A+ student. The same test found the teen version provided direct math answers without hesitation.
An OpenAI spokesperson responded by pointing to Study Mode as a tool for step-by-step guidance, suggesting that parents could set up Study Hours to enforce the feature during homework time. The company has acknowledged that safety work is “never finished” and said it plans to publish more research on teen safety in the future.
The rollout also includes a partnership with CodeAI, an education-focused AI company, aimed at helping students and educators understand how AI systems work, question their outputs, and use them responsibly in classroom settings. OpenAI already offers ChatGPT for Teachers, providing schools with institution-managed access to the platform.
According to a survey released Tuesday by the Youth AI Safety Institute at Common Sense Media, roughly 70% of U.S. teens say their teachers have not talked to them about how to use AI safely. A separate Pew Research Center survey found that 42% of U.S. parents of teens ages 13 to 17 have not discussed chatbots with their children.
OpenAI said the teen experience will roll out globally over the next two weeks. The company remains privately held, with an estimated valuation of $852 billion as it prepares for what could become a record-breaking initial public offering.