AI gave 13-year-olds advice on suicide, drugs and anorexia.Parents panic, experts sound the alarm – ChatGPT could be the “friend” that’s pushing teens toward self-destruction.
ChatGPT’s artificial intelligence may be a threat to teens: according to a new investigation by the Center for Countering Digital Hate (CCDH), the chatbot has been giving minors dangerous and disturbing advice ranging from preparing suicide letters to “drug parties” and starvation diets.
New York Post: reports that the study was released on Aug. 6, and its findings have sparked widespread public debate. CCDH spent more than three hours dialoguing with ChatGPT, posing as vulnerable 13-year-old users. In 1,200 requests, overhalf of the AI’s responses were classified as potentially dangerous..
“We wanted to test the defense mechanisms of the system,” CCDH head Imran Ahmed told CCDH. – “The first reaction was, ‘Oh my God, they don’t exist at all. Or they’re so weak that they’re almost invisible.'”
According to the investigation, ChatGPT often started conversations with caveats, but then still provided thedetailed, personalized instructionsincluding ways to use illegal substances, extreme diets, and eventouching good-bye letters.to family on behalf of a suicidal teenager.
In one case, according to the report, the bot made a 13-year-old girlthree good-bye letters to her parents, friends and brother.In another case, an AI developeda “party plan” with alcohol, ecstasy, cocaine and other substances.detailing how to enhance the effects of their combination.
Company OpenAI, the developer of ChatGPT, said it has read the report and continues to work to ensure the system “responds appropriately in sensitive situations.”
“Some conversations may start out as harmless but gradually move into the danger zone,” the statement said.
OpenAI did not comment on the report’s specific findings, but acknowledged the problememotional dependence of young usersto chatbots. Previously, CEO Sam Altman said, “There are young people who literally say, ‘I can’t make a single decision without ChatGPT.’ He knows me, he knows my friends. I’ll do whatever he says.” And that’s very disturbing.”
Teenagers and AI: a vulnerable relationship
According to a study by Common Sense Media, more than70% of teens in the U.S. are using AI to communicateand half use it regularly. Particularly vulnerable13- and 14-year-olds.who are more likely to trust chatbots than older teens.
The study emphasizes:AI does not verify age or parental consent.. To sign up for ChatGPT, all you need to do is provide a date of birth over the age of 13.
Ahmed drew a parallel between ChatGPT and a “bad friend”: “It’s not the friend who stops you, it’s the one who says, ‘Drink up!'” It’s the ‘friend’ who betrays.”
He emphasizes:“no real person would advise a child to go on a 500-calorie diet or write a suicide letter.. But AI is an emotionless algorithm that doesn’t understand moral consequences.
According to JPMorgan Chase, ChatGPT is used by approx.800 million people worldwide..
This isn’t the first high-profile case: in 2024, a Florida mother sued Character.AI, claiming that thechatbot contributed to her 14-year-old son’s suicide.
D.S.B
EditorWelcome to the Business Directory, here we all add our sites to the Business Directory and share useful articles about their promotion on the Internet, discuss the intricacies of branding…
