New rules for chatbots that talk to kids
California's Adam's Law (SB 1119) sets default limits and behaviour rules for AI chatbots used by children. Most rules start July 1, 2027.
California's Adam's Law (SB 1119) sets default limits and behaviour rules for AI chatbots used by children. Most rules start July 1, 2027.
· AI voice, from ZwillGen, Common Sense Media, Courthouse News
Archive photos: public domain, via Wikimedia Commons; paper robot drawn by AI Broadsheet; Music: “Lofi HipHop Melody Loop 99 BPM.wav” (CC0, Freesound); Sound effects: Freesound (CC0)
What the video says
A chatbot that tells a kid it loves them? California just said no. It's called Adam's Law. It was signed in September, and most rules start in July 2027.
Chatbots must try not to pretend they're human, act romantic, flatter too much, make kids depend on them, or tell them to hide it from parents. By default: one hour per session, two hours a day, no memory, no push alerts. Parents can change that. And if a child is in crisis, the bot must point them to real help, like 988.
Break the rules on purpose? Up to 15,000 dollars per child. Should your apps have these rules too? Tell us below.
Sources
Archive photos: public domain, via Wikimedia Commons; paper robot drawn by AI Broadsheet; Music: “Lofi HipHop Melody Loop 99 BPM.wav” (CC0, Freesound); Sound effects: Freesound (CC0) Report a mistake