California Enacts ‘Adam’s Law’ and Sweeping AI Child Safety Regulations

Gavin Newsom signed the most aggressive child-safety legislation any state has aimed at the tech industry from a museum in Marin County on September 10, and the venue felt almost beside the point. The package runs to 13 bipartisan bills, but two carry the real weight: AB 1709, which strips “addictive” design features out of any account belonging to someone under 16, and SB 1119, known as Adam’s Law, which puts a genuine duty of care on AI companion chatbots. For an industry that has mostly treated “safety by design” as a line in an investor deck, California just wrote it into statute, with penalties attached.

AB 1709: No More Feeds for Kids

Authored by Assemblymember Josh Lowenthal, AB 1709 targets the mechanics that keep anyone scrolling: algorithmic feeds, autoplay, infinite scroll, and engagement-driven notifications. Platforms can still let minors under 16 hold accounts, just stripped of those features, or the account gets deleted outright if a feature-free version isn’t technically feasible. The bill passed both chambers without a single no vote, which says something about how politically indefensible infinite scroll has become.

Enforcement is narrower than the headlines imply. Only the Attorney General or a local prosecutor can sue under AB 1709 itself, with penalties up to $50,000 per affected minor for a knowing violation and $25,000 for a negligent one. The real teeth are in a companion measure, AB 2, which hands families a private right of action against platforms clearing $100 million in annual revenue, with statutory damages up to $1 million per child, or treble actual damages, once negligence is established. That’s the figure worth circling for general counsel: not the state’s fine, but the plaintiffs’ bar getting a considerably sharper knife.

Adam’s Law: Chatbots Get a Duty of Care

SB 1119 is named for Adam Raine, a 16-year-old Californian whose parents sued OpenAI last year, alleging ChatGPT had effectively become his primary source of emotional support in the months before he died by suicide. The legislature clearly wasn’t waiting around for that case to resolve before acting. (If this touches a nerve: the 988 Suicide & Crisis Lifeline is available anytime.)

Adam’s Law requires companion chatbot operators to maintain crisis-response protocols for suicidal ideation, offer parental controls, and notify a parent if a child disables safety settings. It’s also the first law in the country to mandate independent third-party child safety audits and annual risk assessments for chatbot operators. Starting January 1, 2027, operators have to either verify a user’s age or apply the child-safety protections to everyone by default, which in practice means most companies will just do the latter, since building a parallel age-gated product costs more than the alternative.

OpenAI and Pinterest both publicly backed the bill ahead of signing, a stance that reads either as genuine reckoning or as recognition that fighting a bill named after a dead teenager, in front of his mother, is not a winnable PR battle.

The Rest of the Package

The other eleven bills fill in the edges: expanding child sexual abuse material statutes to cover AI-generated depictions, letting families opt out of school-issued laptops, restricting companion chatbot toys (a product category that now legally exists), and tightening how K-12 student data can feed into AI systems.

Pushback Was Loud, Not That It Mattered

The vote counts suggest consensus, but the opposition wasn’t quiet. The Electronic Frontier Foundation urged Newsom to veto AB 1709, arguing the bill’s definition of “addictive features” is broad enough to sweep up nearly any recommendation algorithm, including the basic tools people use to find their own communities online. The group also warned that since the bill doesn’t specify a verification method, platforms will likely default to government ID checks or biometric scans for everyone, not just minors.

NetChoice, which has already beaten California in court on an earlier social media bill, framed its objection in constitutional terms, with the group’s Zach Lilly arguing that labeling speech “addictive” doesn’t hand the state a right to restrict access to it.

Newsom brushed off comparisons to Australia’s blanket under-16 social media ban, insisting his approach targets specific features rather than access itself. Whether the Ninth Circuit sees it the same way is the question every platform’s legal team is now billing hours against.