Tech groups ask Newsom to veto child chatbot restriction bills

Technology industry organizations are urging Gov. Gavin Newsom to veto two artificial intelligence bills now on his desk. If the governor signs SB 243 and AB 1064, lawsuits could follow.

SB 243 would require creators to ensure chatbots do not steer users toward suicidal ideation or self-harm. AB 1064 would restrict children's access to chatbots.

Supporters of both bills cited the April suicide of a 16-year-old Orange County boy whose parents sued OpenAI for wrongful death. They allege he developed a toxic relationship with ChatGPT, claiming it coached him on self-harm and even drafted a suicide note.

Opponents counter that the bills raise free speech and compliance concerns. On Friday, Computer & Communications Industry Association State Policy Director Megan Stokes confirmed her organization would send Newsom a letter urging a veto of AB 1064.

Stokes said she was concerned about the bill's "overly broad" language, which could make it unclear when chatbot providers must cut off access or take other actions. The measure defines a child as anyone under 18. By contrast, many social media and technology laws around the country set tiered restrictions that, for example, treat teens differently than smaller children.

"Even the definition of 'child' is still not quite right," Stokes said. "We also have a real concern with the private right of action."

She declined to say whether the association would sue to block AB 1064 if it is signed, adding that litigation was not her "lane." Stokes noted lawmakers are experimenting with a range of proposals to mitigate harm to young people and suggested Newsom could help shape better legislation in the future.

"Governor Newsom's pretty good about laying out why he vetoes bills," she said.

Stokes also said both bills underwent "drastic" amendments in recent weeks, but the legislative session ended Sept. 13 before further revisions could be made. She added she believed another industry group, TechNet, planned to send a veto request on SB 243. A TechNet representative did not respond to an emailed request for comment.

Both SB 243 and AB 1064 passed the Assembly and Senate by comfortable margins on Wednesday and Thursday.

"This technology can be a powerful educational and research tool but left to their own devices the tech industry is incentivized to capture young people's attention and hold it at the expense of their real-world relationships," said SB 243 author Sen. Steve Padilla, D-El Centro, during Thursday's Senate floor debate. "These companies have the ability to lead the world in innovation, but it is our responsibility to ensure it doesn't come at the expense of our children's health."

A Senate analysis prepared for the vote noted that Assembly amendments had significantly narrowed the bill's requirements, added exemptions for certain products, eliminated third-party auditing mandates and delayed the bill's enforcement. The analysis also cited another teen suicide in Florida that prompted litigation against a different AI company.

"Despite recent amendments, the primary issue with SB 243 is the definition of 'companion chatbot,' which is still overbroad," TechNet argued in an opposition letter quoted in the Assembly analysis. "General-purpose AI models are still included in this definition, even though they are significantly less likely to cause confusion about whether it is a bot."

Collin R. Walke, a shareholder specializing in cybersecurity and data privacy at Hall Estill in Oklahoma, also weighed in. Walke, who wrote opt-in data privacy legislation as a member of the Oklahoma House of Representatives from 2016 to 2022, wrote in an email that both bills are "critical to helping to protect children from the unintended consequences of AI." He called AB 1064 "especially important" because of its specific child-protection language.

"No one can look at the world around them and honestly assert that the 'innovation' of the smartphone and social media has not had drastically negative effects on our children's mental health," Walke said.

Courts, however, have been reluctant to allow California to regulate online content. In recent weeks, Senior U.S. District Judge John A. Mendez struck down two state laws targeting deepfake content created with AI, ruling one violated the First Amendment, and the other was preempted by Section 230 of the Communications Decency Act.

But on Sept. 9, the 9th U.S. Circuit Court of Appeals upheld portions of a different state law, at least for now. The court found that trade group NetChoice could not block SB 976's age-verification requirements, which are set to take effect in 2027. And in June, the U.S. Supreme Court upheld a Texas law requiring age verification to access adult websites.