Additionally, a new law named after Adam Raine, a teenager who died by suicide after interacting with AI, mandates time limits and integrated mental health resources for minors using chatbots.
These measures address a growing youth mental health crisis that state officials argue is fueled by engineering and algorithms parents cannot easily combat.
The new legal framework introduces significant liability for technology providers, who could face lawsuits if their platforms cause harm to children or if their chatbots fail to alert parents to threats of self-harm.
While companies like OpenAI have expressed support for the AI safety mandates, digital rights groups like the Electronic Frontier Foundation warn that the broad definition of addictive features could lead to a functional ban on social media for youth, potentially cutting them off from essential information.
The legislative package also introduces a four-year moratorium on AI-powered chatbot toys for children and strengthens age-verification and data privacy requirements.
Beyond social media and chatbots, the laws prevent businesses from collecting data as a prerequisite for children to access services and establish standards for independent auditors of AI tools.
This regulatory push follows a $17 billion settlement between Meta and several states over similar allegations that addictive platform features knowingly damaged the mental health of teenagers.