Citing recent high-profile fatalities linked to the technology, a prominent California Democrat senator delivered a harsh critique on Wednesday of what she claimed were too loose safety regulations for children using AI chatbots.
Referring to a lawsuit regarding the death of a 16-year-old kid who requested comments from OpenAI’s ChatGPT, Assemblymember Rebecca Bauer-Kahan told POLITICO, “We see reports, like we saw this week, of a young man taking his own life after ChatGPT coached him on how to do it.”
“We’re progressing too quickly with our kids. AI experiments won’t be conducted there.”
The worries expressed by Bauer-Kahan on Wednesday during an AI regulatory panel at POLITICO’s “The California Agenda: Sacramento Summit” coincide with accusations from both Democrats and Republicans that AI businesses are not doing enough to protect children when they utilize so-called “companion chatbots.”
Following the complaint alleging wrongful death, OpenAI said that it was developing an upgrade to enhance its models’ ability to identify and react to indications of mental and emotional trouble.
Several members of Congress have called for inquiries into Meta after it was revealed that the company’s policies permitted chatbots to engage in “sensual” dialogue with children. Additionally, over 40 state attorneys general, including Rob Bonta of California, wrote letters to a number of AI firms on Monday in response to the disclosures of Meta’s chatbot policy and the warning that “if you knowingly harm kids, you will answer for it.”
According to Meta, such discussions with youngsters should never have been allowed, and the company is updating its policies.
Democrats in California are advancing two legislation that would shield children from the possible risks associated with companion chatbots. According to Bauer-Kahan’s original draft, businesses would be prohibited from providing children with “emotionally manipulative” chatbots.
More specifically, Democratic state senator Steve Padilla’s second bill would compel obligatory reporting when people talk about self-harm with chatbots and outlaw addictive compensation systems that are occasionally employed to encourage user interaction with the bots.
On Wednesday, Bauer-Kahan forecast that Governor Gavin Newsom “will sign a lot of AI legislation.”
A ban on state AI regulation, which Congressional Republicans failed to get via their domestic policy megabill earlier this year, is one issue that may persist past Newsom’s term limit. Sen. Ted Cruz is among those who are already keen to bring it back.
The head of the Carnegie Endowment for International Peace and a former justice of the California Supreme Court, Mariano-Florentino “Tino” Cuéllar, stated on Wednesday that the next governor will need to manage it skillfully.
State lawmakers get a good start since they brainstormed methods to avoid federal preemption before the bill was repealed.
