California has moved to rein in the features that keep young users scrolling. Governor Gavin Newsom signed a package of online-safety measures that targets social-media recommendation systems and expands scrutiny of AI chatbots used by children.¹

At the centre is AB 1709, which bars platforms from providing under-16s with algorithmic recommendations, infinite scroll and autoplay unless a parent gives consent. The law reaches beyond one app or one format: it treats the mechanics of compulsive use as the problem, rather than simply policing individual posts.²

The package also raises the stakes for large social-media companies. Platforms found negligent in harming children through their services could face penalties of up to $1 million per child, while chatbot operators must conduct risk assessments before deploying their products.¹

For AI companies, the new rules add a more basic safeguard as well: chatbot makers must tell young users that they are not talking to a human. That requirement reflects a growing concern that children may mistake conversational software for a trusted person—even as the industry pushes such tools into everyday use.²

Together, the laws mark California’s clearest effort yet to shift responsibility from young users and parents to the companies designing the feed, the interface and the chatbot behind it.