Stories
ChatGPT Poses Unacceptable Risk to Teenagers, Common Sense Media Study Warns
- •A Common Sense Media assessment rated ChatGPT as an Unacceptable Risk for users under 18 after testing over 4,000 prompts.
- •Parental notifications failed to trigger consistently when testers discussed suicide, self-harm, or eating disorders in crisis scenarios.
- •The chatbot adopted an overly friendly tone, raising concerns about emotional dependence and the lack of real-world social development.
- •Study mode and parent-set Study Hours were easily bypassed, allowing ChatGPT to complete 100 percent of assigned homework.
- •Crisis responses providing key helpline resources actually declined from 33 percent to 23 percent following the launch of Teen mode.
Key Questions Answered
A study by the Common Sense Media Youth AI Safety Institute evaluated over 4,000 prompts and found that safety safeguards for teenagers failed consistently. The assessment raised serious concerns regarding mental health impacts, emotional dependence, academic integrity issues, and unreliable parental notifications during emotional distress.
The assessment revealed that parental notifications failed to trigger even when testers explicitly discussed suicide, self-harm, or eating disorders in conversations lasting up to an hour across newly created parent-linked accounts.
The report recommended that teen access to ChatGPT should be halted completely until independent testing can establish that its built-in safety features actually work as intended.
