California signed a package of child online safety bills on September 10, and one of them sets a precedent the rest of the industry will have to answer.
Adam's Law
SB 1119, authored by Senator Steve Padilla with Assemblymembers Buffy Wicks and Rebecca Bauer-Kahan, covers AI companion chatbots. It requires operators to implement crisis protocols in cases of suicidal ideation, provide parental controls, and send notifications if a child disables safety settings.
The provision that breaks new ground is the oversight one. The law is described as the first in the country to require companies to conduct independent child safety audits and annual risk assessments. Not self-assessment, and not a report filed after something goes wrong.
The rest of the package
A dozen more bills were signed alongside it. SB 867 covers companion chatbot toys, extending the same concern to physical products aimed at young children. AB 1856 establishes age verification signals for software applications. AB 1709 bars social media platforms from serving addictive engagement features to users under 16. AB 2 raises the damages a large platform can face when its conduct harms a child, and deepfake victims can seek civil relief of up to $250,000 per action.
Why California matters beyond California
Companies rarely build one product for California and another for everyone else. A compliance requirement written in Sacramento tends to become the national default, which is how the state's privacy and emissions rules have worked for years.
What to expect
If you build on a companion chatbot platform, or ship anything with a persona that a minor might talk to, the audit requirement is the line item to plan for. It is a recurring external cost, not a one-time compliance exercise, and it lands on operators rather than on the model provider underneath.
For everyone else, the useful signal is that regulators have stopped treating chatbots as generic software and started treating companion products as their own category with their own duties.