OpenAI launches ChatGPT for Teens with study safeguards

A homework shortcut now comes with a detour. OpenAI has launched ChatGPT for Teens, a dedicated mode for users aged 13 to 17 that sends suspected assignment shortcuts toward Study Mode instead of simply returning an answer. The San Francisco company says the experience is designed to help teenagers learn, think critically and use AI with confidence.

The mechanism is deliberately slower. Study Mode uses guiding questions and step-by-step support to help a student work through a problem, while quizzes and Learning Visualizations are intended to make difficult concepts easier to explore. A new Responsible Homework Reminder can recognize when a teenager appears to be trying to skip the work and redirect them; users or parents can also set Study Hours, which turn Study Mode on at chosen times.

Safety protections are enabled by default. The teen experience places tighter limits around self-harm, suicide, eating disorders, dangerous behavior, graphic violence and sexual or romantic conversations. It also warns users before sensitive image uploads and is designed not to imply that ChatGPT has personal feelings, consciousness or a romantic bond with the teenager. Allison Mishkin, OpenAI’s head of child development, said the company examined cues that could encourage a teenager to form a relationship with the chatbot.

Parents can link accounts to set Quiet Hours—periods when ChatGPT cannot be accessed—and receive notifications in limited high-risk situations, including possible self-harm; OpenAI says eating-disorder notifications are being added. The controls are optional, however. Users who identify themselves as minors, or whom the system estimates to be under 18, are automatically placed in the teen experience, while OpenAI does not verify ages directly, according to Phys.org.

So what changes in practice? A teenager can still use ChatGPT for questions and schoolwork, but the default interaction is meant to put guardrails around two of the most contentious uses: outsourcing assignments and treating an AI system like an emotionally present companion. Parents get scheduling and alerts, while students get a built-in path toward guided practice rather than instant answers.

That promise has a test ahead. TechCrunch reports that it is not yet clear how difficult the new protections will be to evade, since teenagers often find ways around parental controls and other restrictions. OpenAI’s launch follows public scrutiny and lawsuits over chatbot safety, but the product’s success will turn on how its safeguards perform in real conversations—not only on the settings shown at launch.