Skip to contentNewskeryx

21 August 2026

Watchdog says OpenAI's teen ChatGPT safety features often fail

First reported

BBC News, CNBC and 1 other ran this on , 3 days before the next source picked it up.

  • Common Sense Media, a children's digital safety group, tested OpenAI's ChatGPT for Teens and called it an unacceptable risk for young people.
  • Its testing found parental alerts for suicide, self-harm or eating disorder conversations failed to trigger on new accounts, though explicit role-play was refused.

Reported by Fast Company, BBC News, The New York Times