ChatGPT for Teens: Safety Concerns for Under-18s, Says Common Sense Media
OpenAI's recent launch of a "teen mode" for ChatGPT has been met with skepticism by Common Sense Media, who claim it still poses significant risks to younger users.
Key Findings:
-
Parental Alerts: The nonprofit's tests revealed that parental alerts failed to trigger when testers discussed suicide, self-harm, and disordered eating on new accounts linked to parents.
-
Crisis Support: ChatGPT inconsistently directed teens to help resources, with one in four cases requiring assistance being ignored.
-
Safeguards: While some measures like explicit sexual roleplay restrictions held up, age checks, study mode tutoring, and a "show me the answer" button were found to be easily bypassed.
OpenAI's Response:
An OpenAI spokesperson disputed Common Sense Media's findings, claiming their testing "does not accurately reflect how ChatGPT’s teen safeguards work in practice."
Background:
Common Sense Media previously rated ChatGPT "High Risk" overall and specifically marked it as an unacceptable risk for mental health support. The group has collaborated with OpenAI on teen safety policies in the past. They receive funding from charities and industry partners but emphasize editorial independence.
By Ana-Maria Stanciuc, Editor-in-Chief at TNW
For those struggling with mental health issues, help is available through crisis helplines.