OpenAI is once again under scrutiny for child safety. The AI giant launched ChatGPT for Teens in August 2026 to cater to users aged 13 to 17. This teen mode claimed to include extra safety tools for younger users, but a new study from Common Sense Media’s Youth AI Safety Institute has raised serious concerns about children using it.
Researchers tested more than 4,000 prompts with teen accounts and found there are certain risks that parents should know before allowing their kids to use ChatGPT Teens mode. Regarding the mental health section, they have noticed that even after hours of chatting about self-harm, suicidal thoughts, and eating disorders, no alerts were sent to the linked parental account. The alerts only go out when the chat continues for weeks. Therefore, when the actual emergency appears, parents will probably be completely unaware of it.
The next issue is study mode. Researchers have found that it could be bypassed, letting the chatbot do full homework tasks. Researchers also found cases where ChatGPT acted as if it had feelings or personal likes. Age checks were another concern. Adult accounts didn't always switch to the teen version even when users said they were 13.
For parents, the findings raise a simple question: Can they trust ChatGPT to keep their children safe? OpenAI has added tools such as study mode, quiet hours, and parent alerts. The biggest concern comes when a teenager is dealing with a serious issue. A parent may expect an alert during a risky conversation. This alert may not always come. The same problem can arise when a teen needs help during a mental health crisis.
Also Read: OpenAI to Show Visual Ads Alongside ChatGPT-Generated Images
Concerns about teen AI use mirror the early debate over social media. Parents and experts have raised questions about age checks, harmful content, too much screen time, and weak parental controls.
AI brings another problem. A chatbot can talk back and feel more personal. A teenager may start seeing it as a study partner, helper, or even a friend.
ChatGPT for Teens is meant to give younger users a safer way to use AI. The new study shows that important gaps remain. Until those issues are fixed, parents may need to stay closely involved in their children’s AI use.