On 13 October 2025, the Leading Ethical Artificial Intelligence (AI) Development for Kids Act was vetoed by the Governor of California. The Act applied to operators of companion chatbots that simulated sustained humanlike relationships with users under 18. It imposed a ban on systems that encouraged self-harm, suicide, violence, substance use, disordered eating, illegal activity, or sexual interactions with children. It also prohibited chatbots from offering unsupervised therapy, discouraging professional help, or prioritising validation over safety. Violations carried civil penalties of USD 25,000 per case.
Original source