A bipartisan group of U.S. senators, including Richard Blumenthal and Chris Murphy (D-CT), is seeking to ban AI chatbots for children.
At a recent Senate hearing, and again at a press conference announcing the legislation this week, parents of kids who died by suicide said the chatbots were exposing their children to sexual and suicide-related content.
Maria Raine lost her 16-year-old son, Adam, to suicide in April. She blames ChatGPT.
“My son, Adam, ended his life in April after ChatGPT coached him to suicide over the course of months. I'm here today with my husband Matt to support this critical legislation,” Raine said. “It was only after Adam died that we learned what Chat GPT had done to him, and now we know that OpenAI twice downgraded its safety guardrails in the months leading up to my son's death, which we believe they did to keep people talking to Chat GPT.
“If it weren't for their choice to change a few lines of code, Adam would be alive today,” Raine said.
The GUARD Act would require age verification for users. It would also require a disclosure that explicitly told users the conversation they’re having is not with a human.
“1/3 of children today report having deep, intimate relationships with these chatbots,” Murphy said. “These stories that you're hearing of children taking their lives or engaging in life-altering self-harm, this is not the exception. This is becoming a norm, and if we don't do something, it is going to get worse.”
The GUARD Act was introduced by Blumenthal and Sen. Josh Hawley (R-MO). It’s co-sponsored by Murphy and Sen. Katie Britt (R-AL).
When asked why Congress hasn’t acted on the issue yet, Hawley blamed the technology lobby.
“They're the most powerful companies on this hill,” Hawley said. “They spend like nobody's business. And I've said before, joking, but really it's more ruefully, there ought to be a sign outside of the Senate chamber that says, ‘bought and paid for by big tech.’”
One tech industry trade group accused the bill of being too heavy-handed, raising privacy concerns.
“We all want to keep kids safe, but the answer is balance, not bans,” Chamber of Progress K.J. Bagchi with the Chamber of Progress said. “It’s better to focus on transparency when kids chat with AI, curbs on manipulative design, and reporting when sensitive issues arise.”
The legislation comes as Blumenthal continues to push for passage of his Kids Online Safety Act. That bill, which would impose a duty of care on social media sites and require companies to allow users to opt out of algorithms, passed the Senate with an overwhelming majority. It did not get called for a vote in the House.
KOSA critics have raised concerns over potential censorship.
