Regulating the Relationship: What the First Sui Generis Companion Chatbot Statutes Achieve for Minors, and What They Leave Undone

Abstract

AI companion chatbots support extended personal conversations, and many adolescents use them for companionship and emotional support. Two American lawsuits allege that prolonged interactions with these systems contributed to minors’ suicides. In 2025, California, New York and Utah enacted technology-specific statutes addressing companion or mental health chatbots. A close reading of these statutes revisits a longstanding question: should emerging technologies be governed by specialised legislation or general law? The principal contribution of the California and New York statutes is definitional: they recognise the sustained human–machine relationship as an object of regulation. Their substantive duties, however, focus on disclosure and crisis protocols, leaving the design choices that encourage emotional attachment largely unaddressed. Disclosure alone offers limited protection because users can respond socially to machines even when they know they are artificial. The analysis therefore supports a combination of approaches: technology-specific definitions that identify the relationship to be regulated, coupled with duties enforced through established consumer protection mechanisms. This combination provides a basis for regulation, but the enacted duties offer minors limited protection against the commercial design of attachment.