California is now the most aggressive state in the country when it comes to regulating what AI chatbots and social media platforms can do with users under 16. Governor Gavin Newsom signed a package of bills this week that sets hard requirements for AI companies, restricts social media features for teens, and expands what counts as child sexual exploitation in the age of generative AI. These are not vague policy statements. They carry legal liability.
The AI chatbot rules are the most technically specific. Companies will need to implement time limits on how long minors can use their products, surface mental health resources within the product itself, and put safety protocols in place for conversations involving self-harm. If a child turns off safety settings, parents get notified. AI companies will also face mandatory independent child safety audits and annual risk assessments. Fail to comply and you’re looking at legal exposure. This follows a separate set of measures Newsom signed the day before, which created broader audit frameworks for AI companies operating under California law.
The social media piece is where things get contentious. The state now bans platforms from offering users under 16 access to what the law calls “addictive features.” That definition is wide. It covers autoplay video, algorithmic content feeds, and push notifications. For anyone building consumer social products, that’s not a minor carveout. That’s most of the product.
The Electronic Frontier Foundation came out hard against this measure, calling it a functional ban on social media for teenagers. EFF Associate Director of State Affairs Rindala Alajaji argued that restricting minors’ access to digital forums doesn’t make them safer, it just limits their participation. The organization also flagged a problem that critics have raised with similar age-verification laws before: enforcing age restrictions requires collecting more personal data from all users, not just minors, which creates new privacy risks at scale.
Still, the broader package includes provisions that are harder to argue against. Newsom also signed bills that:
- Expand the legal definition of child sexual exploitation to include AI-generated or digitally altered images
- Restrict targeted advertising aimed at minors
- Regulate how AI systems handle data collected from K-12 students
The bigger picture here is that California is building a regulatory stack for AI and youth safety faster than federal lawmakers are moving. Companies like Meta, Snap, Google, and any startup running an AI chatbot with a consumer audience now have to treat California compliance as a serious product and legal priority. Given that California accounts for a significant share of the U.S. tech market, what gets built for California often gets built everywhere. That’s the real weight of what Newsom signed this week.
