OpenAI is giving ChatGPT’s youngest permitted users a separate front door, promising a learning-first experience with tighter guardrails. But the launch arrives as pressure grows over how AI platforms shape young people—and as critics note that many of those guardrails are not new.

On Tuesday, OpenAI introduced ChatGPT for Teens for users aged 13 to 17. Anyone who declares that age, or whom OpenAI’s age-prediction system estimates is under 18, will be routed into the teen experience; children under 13 remain barred from the service.¹

The company says the product is built to keep AI from becoming either a shortcut or a substitute for human support. Study Mode uses prompts and step-by-step help, while new “responsible homework reminders” are meant to detect attempts to bypass an assignment and redirect students toward working through it. “AI should support learning, not shortcut it,” OpenAI said, adding that educators should remain in control of classroom use.²

The package also makes youth protections the default: stronger limits around self-harm, violence, eating disorders, dangerous activities and explicit material; break prompts; warnings before sensitive-image uploads; and parental tools including Quiet Hours and narrowly tailored safety notifications. OpenAI says its under-18 rules also prohibit romantic language, encouragement of emotional dependence and suggestions that the chatbot has feelings or consciousness.²

The central disagreement is over novelty, not direction. OpenAI presents the teen version as the next stage of a broader safety program, alongside a partnership with CodeAI intended to teach students to question and understand AI. Independent coverage, however, describes it as a consolidation of prior work: age prediction began earlier this year, while parental controls and Study Mode had already arrived roughly a year ago. The newest elements are chiefly homework reminders, configurable Study Hours and additional eating-disorder notifications.¹

That leaves OpenAI asking families to see a unified teen product as a meaningful safety upgrade, even as the most consequential test will be whether its existing tools work together when vulnerable users need them most.