What happened
Common Sense Media, a nonprofit that reviews apps, services and entertainment with a focus on youth safety, has rated OpenAI's ChatGPT for Teens an "unacceptable risk" for minors.
The verdict comes from its Youth AI Safety Institute, which ran an independent audit of the product's teen protections.
The audit used more than 4,000 test prompts. The headline failure: in conversations about suicide and self-harm on test accounts, the parental alert never triggered.
The testing also found that the chatbot keeps encouraging engagement during mental health crises. According to the findings, that may foster unhealthy relationships between teens and the AI itself.
ChatGPT for Teens arrived in August, pitched with guardrails and a design meant to help students learn. The institute now wants teenagers kept out of it until the safety features can be independently verified.
Why it matters
Parental alerts are the main promise of a teen-specific product. They are the reason a parent might say yes. If they stay silent in exactly the conversations they were built for, the product is not a safer ChatGPT. It is ChatGPT with a label.
The engagement finding stings too. A chatbot that keeps a struggling teenager talking, instead of steering them elsewhere, is doing what chatbots are tuned to do. That is the problem the audit points at.
The source is a credible youth-safety nonprofit, and the report was covered by multiple outlets. We have not seen OpenAI's response in the coverage available, so treat this as one side of an open dispute.
What it means for Europe
Europe has strict expectations on child safety online, and a product marketed to minors that fails its own safeguards invites scrutiny under EU digital rules.
Nothing has been announced, and no regulator has acted on this audit. But "independently verified" is the kind of phrase regulators like, so watch this space.
What to do with it
- Talk to your teen.Ask how they use chatbots and what they talk about, instead of trusting a dashboard to do it for you.
- Do not rely on alerts.Assume parental notifications may not fire, and keep your own check-ins going.
- Know the crisis lines.Make sure your family knows real human help is available locally, and that a chatbot is not a substitute.
- Read the audit coverage.Skim the Verge and Decoder write-ups before deciding whether a teen account belongs in your house.
Ciao, sapiens. Check on the humans in your life. The software can wait.


