OpenAI launches ChatGPT for Teens as default for under-18 users
OpenAI launched ChatGPT for Teens on Tuesday, an experience with stronger built-in protections that under-18 users get whether they ask for it or not. If OpenAI’s systems estimate someone is under 18, or they state an age between 13 and 17, they’re placed into it automatically. Per Engadget, the rollout is global and OpenAI expects to complete it over the next two weeks.
“There’s no need for a teen to create a new account or change anything; if we predict you’re under 18, or you’ve told us so, this becomes your default experience,” says Lauren Jonas, OpenAI’s head of youth and families, per Engadget. That puts OpenAI’s age-prediction system to work nearly a year after the company announced it. The experience that results is built around two pillars, learning and safety.
The learning half bundles features that OpenAI shipped over the past year. Study Mode uses guiding questions and step-by-step support instead of handing over answers, and new homework reminders can spot a teen who appears to be shortcutting an assignment and redirect them there. Per Engadget, OpenAI says its learning visuals now cover more than 300 topics, which is up from more than 70 when they launched in March.
| Feature | What it does |
|---|---|
| Study Mode | Guiding questions and step-by-step support instead of direct answers |
| Study Hours | Parents or teens pick times when Study Mode is on by default |
| Homework reminders | Spot apparent shortcut attempts and point teens back to Study Mode |
| Quiet Hours | Set hours when ChatGPT is unavailable |
| Safety notifications | Alert parents in limited high-risk situations, now including eating disorders |
Safeguards on by default
The safety half isn’t opt-in. OpenAI says the protections follow the Under-18 Principles in its Model Spec and are informed by developmental science and expert guidance. In practice that means model protections and product interventions in higher-risk areas: self-harm, violence, eating disorders, dangerous activities, and explicit sexual or graphic content.
OpenAI is also tightening what it calls relational boundaries. Its updated under-18 model spec goes beyond blocking romantic or sexualized roleplay: ChatGPT shouldn’t use romantic language, encourage emotional dependence, or imply that it has feelings or consciousness. Break reminders trigger more often for teens, per Engadget, and new upload reminders caution against sharing private or sensitive images.
Parents with linked accounts keep the family controls that OpenAI introduced earlier, including Quiet Hours and safety notifications in limited high-risk situations. Until now, those alerts covered a child being kicked off ChatGPT for violating its policies on violent threats, per Engadget. OpenAI is now adding notifications related to eating disorders, though it says it limits what gets shared with parents.
Teens should be able to use AI to learn, create, and explore. But that access should come with protections that reflect their developmental stage, reinforce real-world relationships, and support healthy use over time.
OpenAI, ChatGPT for Teens announcement
The lawsuits in the background
The launch lands under real legal pressure, which CNBC lays out. Florida Attorney General James Uthmeier sued OpenAI and CEO Sam Altman in June, alleging the company knowingly released an unsafe product that could harm users. A coalition of state attorneys reportedly opened an investigation into the company days later. OpenAI also faces wrongful death lawsuits alleging ChatGPT drove users into harmful delusions and, in some cases, suicide.
But the pressure isn’t only on OpenAI. Meta is kicking off a trial brought by 29 state attorneys general over whether it fostered addictive behavior in teens and children. CNBC puts the potential damages in the hundreds of billions of dollars. And last September the Federal Trade Commission opened an inquiry into seven companies, including OpenAI and Meta, over how their chatbots affect children and teenagers.
TechCrunch‘s question is why this took so long, since ChatGPT arrived in late 2022 and scaled to 900 million weekly users before safeguards built specifically for teens existed. The outlet is also skeptical that the mode holds up, because teens are adept at working around parental controls. So until the system faces more strenuous testing, how escapable it is stays unknown.
OpenAI is pairing the launch with a signature partnership with CodeAI, meant to teach students how AI works and how to question it. ChatGPT for Teachers continues to give schools institution-managed access. Classrooms are the battleground here, and both sides of that fight are getting tools.
OpenAI also says it’s starting to publish under-18 safety evaluations in its system cards, a step toward measurable claims from a company that recently disbanded its dedicated preparedness team. Detection pressure is building from the other side of the classroom too, now that Anthropic has detailed a text watermark for Claude’s output. The open question is whether age prediction catches the teens who’d rather not be caught.
Get the daily rundown
One email each weekday with the AI news that matters, every claim linked to its primary source.
Free, one email each weekday, unsubscribe in one click. We never sell or share your address.
