California Tightens AI Rules as ‘Doomsday’ Warnings Grow

Hands using smartphone with glowing justice scale and AI icons — Photo: Suri_Studio / Shutterstock

California just did something no state legislature had managed before: it wrote enforceable, product-level safety rules into law for both social media design and AI companion chatbots, in the same signing ceremony, with the tech industry’s fingerprints on the final language rather than only its objections.

At a Glance

Gov. Gavin Newsom signed 13 child-safety and tech bills on Sept. 10, 2026, in Sausalito, covering social media design and AI chatbot conduct.
AB 1709 bans “addictive” features — infinite scroll, autoplay — for platform users under 16.
SB 1119, “Adam’s Law,” forces AI companion chatbots to detect self-harm queries, notify parents, and follow crisis protocols.
SB 813 creates the first state-sanctioned system of independent auditors empowered to certify AI systems against California law.
The package follows Newsom’s 2024 veto of a stricter chatbot bill — a reversal driven by mounting litigation, a jury verdict against Meta and YouTube, and a documented wrongful-death case against OpenAI.

What Was Actually Signed

On Sept. 10, 2026, Newsom convened a signing event in Sausalito and enacted a package described by his office as delivering “first-in-the-nation AI safeguards to protect Californians.” The number is consistently reported across outlets as 13 bills, spanning two distinct regulatory targets: the addictive design of social platforms and the behavior of AI companion chatbots. That dual scope matters. Sacramento has regulated data privacy and age verification before; this is the first time it has gone after the mechanics of engagement itself — the scroll, the autoplay, the algorithmic nudge — and paired that with rules governing how a chatbot must respond when a minor types something alarming.

The centerpiece consumer-facing measure is AB 1709, authored by Democratic Assemblymember John Sloan with Republican Assemblymember Josh Hoover as co-author — a bipartisan pairing that undercuts any framing of this as one-party overreach. It prohibits platforms from offering infinite scroll, autoplay video, and similar engagement-maximizing features to users under 16. The mechanism is blunt but precise: rather than policing content, which runs into First Amendment friction, it targets the interface design that keeps a young user locked to a feed past the point of intent.

Adam’s Law and the Chatbot Problem

SB 1119 — Adam’s Law — is the more novel piece of the package, because chatbots occupy a regulatory space with almost no precedent. Named for Adam Raine, a teenager whose family’s wrongful-death litigation against OpenAI surfaced internal chat logs central to the bill’s drafting, the law requires AI companion products to implement parental controls, notify parents when a minor disables safety settings, and follow defined crisis protocols when a user raises the subject of suicide or self-harm. Raine’s mother said publicly the law “will save lives and prevent other serious harms” — a claim that carries real weight given her family supplied the evidentiary record that shaped the statute. Notably, OpenAI itself lobbied Newsom to sign the bill rather than fighting it, a detail that separates this fight from the industry’s usual posture of resisting design mandates outright.

Behind Adam’s Law sits SB 813, which the governor’s office calls a “first-in-the-nation framework for independent verification organizations” empowered to assess AI systems for compliance with state law. This is the piece with the most durable structural consequence. Content rules age quickly as products evolve; an audit infrastructure — third-party bodies with statutory standing to test and certify AI systems — is built to outlast any single chatbot’s feature set. It is the closest analogue California has to financial auditing or vehicle safety certification, applied to a technology barely five years into commercial existence.

Why Newsom Reversed Course

The politically interesting fact is not that Newsom signed these bills — it’s that he vetoed a stricter version of chatbot regulation roughly two years earlier. That earlier veto reflected a governor wary of constraining an industry central to California’s economy and tax base, and confident that lighter-touch rules would suffice. What changed the calculus was not abstract advocacy but concrete legal exposure: a California jury found Meta and YouTube liable for harms tied to addictive design, one of the first verdicts of its kind, and courts in Massachusetts and California began rejecting Section 230 defenses specifically when claims target design choices rather than user content. That erosion of the industry’s most reliable legal shield removed the argument that platforms were untouchable, and gave Sacramento legislative room it didn’t have in 2024.

The Ninth Circuit’s partial rejection of an injunction against California’s earlier Protecting Kids from Social Media Addiction Act reinforced the same signal: courts were increasingly willing to let California’s youth-design rules stand, at least in part, against constitutional challenge. A governor who vetoes into that kind of legal headwind risks looking out of step rather than principled — and Newsom’s reversal reads less as inconsistency than as recalibration to a legal landscape that had genuinely shifted underneath him.

Where This Fits Nationally and Globally

California is not improvising in isolation. Virginia has moved on companion-chatbot bans for minors capable of encouraging self-harm; the European Parliament has floated a harmonized digital age threshold of 16 for platforms including AI companions; and the European Commission has already found Meta preliminarily in breach of the Digital Services Act over the addictive design of Instagram and Facebook. The pattern across jurisdictions is consistent: engagement-optimized design and unsupervised chatbot conversation with minors are being treated as a distinct regulatory category, separate from the older fights over content moderation and data privacy. California’s package is the most legislatively dense version of that pattern to date, but it is a data point in a trend, not an outlier.

▫️Newsom Signs Major AI & Tech Laws as Fears Over AI Risks Intensify▫️

🤖 First-in-the-nation AI oversight: Gov. Gavin Newsom has signed SB 813 and AB 1405, creating a framework for independent assessments of AI systems and a state registry of AI auditors. The goal is to make AI…

— Washington Report (@Washington_Rep) September 11, 2026

What Happens Next

None of this is self-executing. The enrolled bill text, chapter numbers, and operative dates for all 13 measures were not consolidated into a single public record at signing, and enforcement mechanics — how compliance will be measured, how the independent verification bodies under SB 813 will be accredited, how platforms will contest findings — remain to be built out in regulation. Industry pushback, expected largely from Meta and comparable platforms, will likely center on compliance cost and the technical feasibility of the under-16 design restrictions rather than outright rejection, given that OpenAI’s cooperative posture on Adam’s Law has already fractured any unified industry front. The laws are enacted; whether they function as intended is the next chapter, and it will be written in courtrooms and compliance filings rather than press conferences.