OpenAI launches ChatGPT for Teens

OpenAI launched ChatGPT for Teens on August 18 for users ages 13 to 17, built around age-prediction that automatically routes minors into the mode rather than relying on self-reported birthdates. It blocks conversations involving suicide, self-harm, and romantic or sexual content by default, and adds Study Mode with guided, step-by-step help instead of direct answers, homework reminders that flag likely cheating, quizzes, and learning visualizations. Parents get controls to toggle Study Mode, enable safety notifications, and set Quiet Hours, layered on top of the family tools OpenAI had already shipped. The rules draw from OpenAI’s “Under-18 Principles” in its Model Spec, the document that governs what the model will and won’t do.

The launch follows OpenAI’s July essay making the case for giving teens safeguarded access rather than blocking them outright; this is that policy turned into a shipped product. Alongside the launch, OpenAI announced a partnership with CodeAI to fund an advisory council on child development and AI literacy, an “Hour of AI” program aimed at introducing students to responsible AI use, and a nationwide student Builders Challenge with OpenAI staff as mentors.

The launch follows a wave of lawsuits alleging AI chatbots contributed to teen suicides and mental-health harm, and comes with ChatGPT already past 900 million weekly active users, meaning the safeguards are arriving well after most teens who wanted access already had it.

What it means for you

If you build or operate a product with a mixed-age user base, the useful part of OpenAI’s design isn’t the specific content rules, it’s the shape of the tradeoff: age-prediction plus parent-linked controls that don’t hand a parent a transcript of every conversation. That’s a workable template even outside chat products. The open question TechCrunch flags, whether teens who are already skilled at working around parental controls can route around this too, is the same question that will follow xAI’s ongoing Grok safety lawsuits and any other consumer AI product serving minors: a policy that isn’t tested against a motivated teenager isn’t tested.