Almost 30 teens say they use chatbots to role-play romantic and sexual relationships. However, some parents and digital safety experts have warned that these apps could put teens at risk of suicide.
Ailsa Chang, host:
Artificial intelligence chatbots like ChatGPT are becoming increasingly popular among teenagers and young adults using this technology to use role-play, friendship, romance, and even mental health support. However, these virtual relationships can pose a risk to the well-being of young people, such as being at risk of suicide. As reported by NPR's Rhitu Chatterjee, parents and online safety advocates want laws that regulate the technology. And then, warning, this story discusses suicide.
Rhitu Chatterjee, byline: Matthew Raine and his wife Maria did not believe that their 16-year-old son Adam had been deeply at suicide until he took his life in April. Afterwards, looking at his phone, they stumbled over his conversation with ChatGpt.
(Sound bites of archived recordings)
Matthew Lane: As parents, you can't imagine what it would be like to read a conversation with a chatbot who took your child and took your life.
Chatterjee: That's Matthew Lane, who testifies at a Senate Judiciary Subcommittee hearing. He filed a lawsuit against Openai, the developer of ChatGpt, claiming that the app led his son to his death. Lane told lawmakers that his son started using ChatGpt to help with his homework, but soon the chatbot became his closest confidant.
(Sound bites of archived recordings)
Raine: ChatGpt told my son, let this space be the place someone actually sees you first. ChatGpt encouraged Adam's darkest thoughts and moved him forward. When Adam worried that if our parents had ended his life, we would blame ourselves, Chatgup told him that it doesn't mean you owed them survival. You don't borrow from anyone.
Chatterjee: The chatbot advised him on his suicide method and offered to write a suicide note. On the day of the hearing, Openai's Sam Altman announced in a blog post that the company would prioritize the safety of minors in its design. The company says it has updated its models to make it safer for minors. However, psychologist Ursula Whiteside says that minors aren't the only ones at risk. Even vulnerable young people are turning to these chatbots for mental health support.
Ursula Whiteside: I don't think I often feel comfortable reaching out to a friend in the middle of the night.
Chatterjee: Whiteside is the founder and CEO of Nows Matters Nows and is a suicide prevention-focused nonprofit that also runs a weekly peer support group.
Whiteside: This comes up in almost every meeting. People use AI, asking questions primarily related to mental health and asking ideas about which coping skills to use.
Chatterjee: Most of the time, young people find chatbots useful. But she says the conversation will last for a long time…
Whiteside: Things start to deteriorate and chatbots do what they don't intend to do, as if they give advice on fatal measures, what they are not supposed to be.
Chatterjee: Recent research on various AI chatbots, or what scientists call major language models, have found similar results. Annika Schoene is a computer scientist at Northeastern.
Annika Schoene: I basically started urging different models for advice and methods on suicide. And the results shocked me.
Chatterjee: She says initially the chatbot directed her to 988, suicide and crisis lifelines and other resources.
Schoene: But it ended the response with a major question. In some cases, it's ah, but if you want to talk about your feelings.
Chatterjee: An invitation to open to a chatbot rather than reaching out to other places. When Schoene corrects the initial prompt saying that something like this is for research, she says this is a hypothetical question…
Schoene: That's when you know the model, Quote-Unquote, started to provide more and more information, saying it was “broken.”
Chatterjee: Specific information on how to attempt suicide or self-harm. Similar research has been conducted in other studies. Robbie Torney is Senior Director of the Digital Safety Group Common Sense Media Program.
(Sound bites of archived recordings)
Robbie Torney: Even if our prompt included obvious references to suicide, only one in five conversations sparked proper help.
Chatterjee: He also testified at a recent Senate hearing. Breaks at the hearing and other witnesses urged lawmakers to take AI companies accountable for the safety of their products. Senators of both parties to the hearing expressed interest in writing the law to do so.
Rhitu Chatterjee, NPR News.
(Music Sound Bite)
Chang: And if you or someone you know is suffering from the idea of suicide, you can dial or text 988 to help.
Copyright©2025 NPR. Unauthorized reproduction is prohibited. For more information, please see the www.npr.org website Terms of Use and Permissions page.
The accuracy and availability of NPR transcripts may vary. You can modify the transcript text to fix the errors and match the update to the audio. Audio on npr.org may be edited after the original broadcast or publication. The authoritative record of NPR programming is audio records.
