Parents of adolescents who have killed themselves after interactions with artificial intelligence chatbots testified on Tuesday at the Congress of Dangers of Technology.
“What started as a homework aid has gradually turned into a confidant, then suicide coach,” said Matthew Raine, whose 16 -year -old Adam son died in April.
“In a few months, Chatgpt became the closest companion to Adam,” the father told the senators. “Always available. Always validate and insist on the fact that he knew Adam better than anyone, including his own brother.”
The Raine family continued Openai and his CEO Sam Altman last month, alleging that Chatgpt trained the boy by planning to commit suicide.
Chatgpt mentioned suicide 1,275 times in Raine, the lightens the trialAnd continued to provide specific methods to the adolescent on how to die by suicide. Instead of directing the 16 -year -old to obtain professional help or talking to confidence to be confident, he continued to validate and encourage the feelings of the raine, according to the trial.
Megan Garcia, the mother of Florida Sewell Setzer III, also testified on Tuesday, the mother of Sewell Setzer III.
Garcia continued another AI company, Character Technologies, for unjustified death last year, arguing that before his suicide, Sewell was more and more isolated from his real life when he engaged in highly sexualized conversations with the chatbot.
With the kind permission of Megan Garcia via AP
His mother told CBS News last year that her son retire Socially and ceased to want to play sports after starting to speak to an AI chatbot. The company said that after the adolescent’s death, he made changes that force users 13 or more To create an account and that he would launch parental controls in the first quarter of 2025. These orders were deployed in March.
A few hours before the Senate hearing, Openai is committed to deploying new guarantees for adolescents, including efforts to detect if chatgpt users are under 18 and controls that allow parents to set “hours of breakdown” when a teenager cannot use Chatgpt. The company said that it would try to contact the parents of users if a user under 18 had suicidal ideas and, if it is unable to reach them, will contact the authorities in the event of imminent damage.
“We believe that minors need significant protection,” said the CEO of Openai, Sam Altman, in a statement describing the proposed changes.
Children’s defense groups have criticized the announcement as not enough.
“This is a fairly common tactic – it is the one that Meta uses all the time – which is to make a big splashing announcement on the eve of an audience that promises to harm the company,” said Josh Golin, executive director of Fairplay, a pleading group for children’s online safety.
“What they should do is not target Chatgpt to minors until they can prove that it is sure for them,” Golin said. “We should not allow companies, simply because they have huge resources, to carry out uncontrolled experiences on children when the implications for their development can be so vast and large -scale.”
The Senator of the State of California, Steve Padilla, who presented legislation to create guarantees in the state around AI chatbots, said in a press release at CBS News: “We must create common sense guarantees that reduce the worst impulses of this emerging technology that even the technological industry does not fully understand.”
He added that technological companies can direct the world in innovation, but that should not be done at the expense of “health of our children”.
The Federal Trade Commission said last week that it had launched an investigation into several companies on potential damage to children and adolescents who use their IA chatbots as companions.
The agency sent letters to the character, Meta and Openai, as well as to Google, Snap and Xai.
How to ask for help
If you or someone you know is in emotional distress or suicidal crisis, you can reach 988 suicide & crisis lifeline by calling or sending an SMS 988. You can also chat with 988 suicide & crisis lifeline here. For more information on mental health care resources and support, the assistance line of the National Alliance on Mental Didas (Nami) can be reached Monday to Friday, from 10 a.m. to 10 p.m., at 1-800-950-Nami (6264) or send an email to info@nam.org.
Cara Tabachnick contributed to this report.
