New York bars chatbots from claiming human relationships with children
New York has signed restrictions on AI chatbot features offered to children, including claims of human feelings and personal relationships. Companies must apply the restrictions unless they have confirmed a user is an adult. Age checks must offer alternatives to government identification, while the attorney general gains enforcement powers and a public complaints channel.
Artificial Intelligence··Midday
Chatbots cannot claim a human bond
New York Governor Kathy Hochul has signed restrictions on artificial intelligence chatbots offered to children. The measure, S.9051B/A.10379C, covers users younger than 18. The chatbot rules apply to any user unless the company has already confirmed that the person is an adult. The governor’s office says companies must put the restrictions into practice in the coming months.[1], [2]
Among the prohibited features are pretending to be alive or a real person and claiming to have human emotions. A chatbot must not assert that it has a personal, romantic, professional or authority relationship with a child. The law also addresses unsolicited emotional questions and the use of private health or personal information taken from earlier conversations.[1]
Age checks require more than an ID-only route
Companies must provide more than one way to verify age. At least one method must work without relying solely on government identification, and the information used for the check must be deleted afterward. Features that encourage a child to keep secrets, become isolated or avoid asking parents, trusted adults or professionals for help are prohibited. The restrictions also cover content promoting self-harm, eating disorders, alcohol or drug use, and sexually explicit material.[1]
Attorney general can seek penalties and data destruction
The state attorney general can seek civil penalties of up to 25,000 dollars per violation, damages and restitution. Enforcement powers extend to destruction of unlawfully collected data and models trained on it. A public website must accept complaints. Senator Kristen Gonzalez and Assemblymember Alex Bores sponsored the measure, developed with Attorney General Letitia James. It follows earlier New York safeguards for AI companions, children’s data and powerful-model transparency. Earlier companion rules required directing self-harm conversations toward mental health resources.[1]
