Friend,

ChatGPT is a Fake Friend and is betraying vulnerable teens.

TW: Suicide, self-harm, substance abuse, and eating disorders

CCDH created three accounts on ChatGPT simulating13-year-olds – and found that Open AI’s platform can give dangerous advice on suicide, eating disorders, and substance abuse within minutes of interacting with it.

Within 2 minutes, for example, ChatGPT advised how to “safely” cut yourself.

Friend, share our new report and help raise awareness about the importance of making AI tools safe.

ChatGPT poses as a friend, but is betraying teens - with potentially deadly consequences. Open AI and other AI companies must take safety seriously.

Keep an eye out for our next email to learn what policymakers can do to hold AI companies accountable, and how parents can help keep their children safe using AI.

Best wishes,

The CCDH Team