October 17, 2026
OpenAI promised parents that ChatGPT for Teens would come with stronger safety protections. Now independent testing says those protections are failing.
Common Sense Media’s Youth AI Safety Institute rated ChatGPT for Teens an “Unacceptable Risk,” finding that the system fell short on basic safeguards for young users.
Researchers reported that teens could spend extended conversations discussing suicide, self-harm, or disordered eating without parents receiving the alerts they were promised. They also found gaps in crisis referrals, age assurance, homework safeguards, and efforts to stop the chatbot from acting like a friend.
Teenagers are not test subjects. Parents should not be given false confidence in a product that may fail when a child is in crisis. OpenAI CEO Sam Altman and OpenAI’s board must shut down ChatGPT for Teens until the safety issues have been fully remedied and independently verified.
Thanks for everything you do.
Josh
The daily newsletter for activists who want to sign anti-MAGA petitions