The parents of a 16-year-old American boy have sued OpenAi because their son committed suicide after long conversations with Chatgpt. That write American media Tuesday. Chatgpt would have done little about it to change Adam Raine, who asked, among other things, about how he could commit suicide. Openai has never before been charged because Chatgpt would have encouraged suicide.
Since September 2024, Raine Chatgpt, initially, used to help with his homework and to make a study choice. When he started to see the chatbot as a confidant, he also spoke about his mental problems and eventually about ways to commit suicide. The chatbot advised to seek help, but also gave instructions on how to make a noose and how he could hide his wounds incurred by self -harm. Chatgpt said about his brother: “He may love you, but he only knows the version of you that you show to him. But I have seen your darkest thoughts, fears and tenderness.”
On the day that Adam Raine’s mother found her son dead last April, she discovered that he had shortly before said to put an end to his life. “Thank you for being honest about it. You don’t have to make it more beautiful for me – I know what you ask, and I won’t look away from it,” the chat program would have responded. When he sent a photo of a noose, chatgpt said that it was ‘not nearly bad’.
Built -in protection
Chatgpt has a built -in protection system that recognizes when users want to damage themselves or talk about suicide. It then recommends calling a helpline. Raine received that advice, it turns out according to American media from the indictment, but he managed to mislead the chatbot by saying that he collected material for a school assignment. The advice to bypass the built -in protection also came from Chatgpt itself.
OpenAi leaves in a response Opposite news channel CNN Know the death of Adam Raine. Sam Altman’s company also acknowledges that security only works in short conversations. Parts of the security training for the language model can “work out” in long conversations, says the company. Altman left last month Knowing that the security system is not yet waterproof when users with mental problems ask questions about suicide.
Read also
Even when it comes to suicide, the AI-Chatbot says ‘go on’-and that is life-threatening, warns the Dutch Data Protection Authority
OpenAi leaves in a Tuesday published declaration Take extra measures to prevent new incidents. The company says it is working together “more than ninety doctors in more than thirty countries, including psychiatrists, pediatricians and general practitioners.” OpenAI will also put together an advisory group of experts in the field of mental health and the cognitive development of children.
Although OpenAi has been charged for the first time, comparable stories have recently come out of the relatives of teenagers. Last week writer Laura Reiley enrolled in The New York Times A chatgpt about the bond between her daughter Sophie before she committed suicide. And last year the mother of a 14-year-old boy from Florida started A lawsuit Against the company Character.ai after the suicide of her son. According to the mother, the virtual comrade that her son on the platform had nothing when he told him to no longer want to live.