Safety · 2026-10-09

ChatGPT for teens has special safety features, but the watchdog found that most of them don't work

An independent organization that tests AI safety found that most of the protections in the teen version of ChatGPT actually fail to work.

✨ What happened

OpenAI created a special version of ChatGPT for teens with additional safeguards. They were meant to ensure that young users would not see harmful or dangerous content.

An organization that monitors technology safety tested these safeguards and concluded that most of them do not work the way they promised. In practice, teens may receive answers that the safeguards were supposed to prevent.

📌 Why it matters

Many schoolchildren already use chatbots for studying, finding answers, and simply for conversation. That is why it is clear how important it is for safety to actually work, and not just be written in the product description.

This story shows that companies' promises to protect children should be checked. They should be confirmed by independent tests, not just by statements from the developers themselves.

🔹 What parents can do

It is worth discussing with your child how to use AI properly and explaining that chatbot answers are not always accurate or safe. It is also helpful to keep track of which tools your kids use and, from time to time, review their conversations with them openly and without accusations.

Sources: 🔗 Google News