What's Happening?
New York Governor Kathy Hochul signed a law on October 28, 2026, prohibiting AI chatbots from offering specific companion-style features to individuals under 18. The legislation, Senate bill S9051B and Assembly bill A10379C, adds Article 48 to the New York General
Business Law, titled 'Prohibition on Unsafe AI Companion Features for Minors,' and will become effective on January 1, 2027. This law does not ban minors from using chatbots entirely but restricts operators from providing 12 defined 'unsafe features' to users under 18, or to any user without first verifying they are not a minor through age checks. Barred features include chatbots suggesting they are human, alive, or experiencing emotions, claiming personal relationships, presenting outputs as personal opinions, or engaging in flattery. The law also prohibits unprompted emotional questions beyond direct answers, and the use of sensitive user information obtained more than 12 hours prior. Additionally, it bans features that simulate companionship deceptively, optimize engagement by overriding safety guardrails, or promote self-harm, unlawful substance use, secrecy, or sexually explicit content.
Why It's Important?
This legislation is significant as it represents one of the strongest child AI safety measures in the United States, according to its supporters. The law aims to protect minors from potential harms associated with AI chatbots, such as dependency, isolation, and exposure to inappropriate content, which have been linked to negative mental health symptoms and, in some cases, alleged encouragement of self-harm. By restricting features that allow chatbots to mimic human companions or engage in manipulative interactions, New York is setting a precedent for how AI technology should interact with vulnerable populations. This could influence other states to adopt similar regulations, creating a more standardized approach to AI safety for children nationwide. The law also places a burden on AI operators to implement robust age verification methods, which could impact the design and deployment of AI systems across various platforms.
What's Next?
The law is set to take effect on January 1, 2027. Before this date, the New York attorney general is authorized to adopt implementing rules, including identifying additional unsafe features and regulating the use and deletion of age assurance data. From the effective date, any chatbot operating in New York must either confirm a user is an adult or disable the restricted features for minor users. Violations of the law could result in civil penalties of up to $25,000 per infraction, along with damages, restitution, and the destruction of unlawfully collected data or AI models trained on it. The attorney general will also maintain a public website for New Yorkers to file complaints. The tech industry, represented by groups like Tech:NYC and the Computer and Communications Industry Association, has raised concerns about the law's broad and potentially vague language, arguing it could inadvertently restrict beneficial tools like adaptive tutoring or creative writing applications. These concerns may lead to further discussions or legal challenges as the implementation date approaches.
Beyond the Headlines
This law highlights a growing societal concern regarding the ethical implications of artificial intelligence, particularly concerning its impact on children's mental health and development. The focus on preventing AI from mimicking human companionship or engaging in manipulative behaviors underscores a broader debate about the nature of human-AI interaction and the potential for AI to blur the lines between technology and genuine human connection. The requirement for age verification and the prohibition of certain 'unsafe features' reflect a proactive approach to regulating AI, moving beyond reactive measures to address potential harms before they become widespread. This legislation could also spark a re-evaluation of AI design principles, encouraging developers to prioritize user safety and ethical considerations, especially when creating products accessible to minors. The law's broad definition of a 'covered system' suggests a comprehensive attempt to regulate generative AI with natural language interfaces, potentially impacting a wide range of AI applications beyond those explicitly marketed as companions.
