New York signs AI companion restrictions for minors, effective in 2027

What happened?

Governor Kathy Hochul signed S9051B on October 28, 2026, as Chapter 368, restricting specified unsafe AI companion features for minors in New York. The active B text makes the main law effective January 1, 2027; it authorizes preparatory rulemaking immediately. This is distinct from the state’s earlier AI companion disclosure law and is not a blanket ban on chatbots.

The statute covers generative AI systems with a natural-language interface that provide ongoing adaptive responses, including but not limited to products marketed as companions. Listed features include pretending to be human or to have a personal relationship, certain uses of earlier sensitive personal information, and outputs promoting self-harm, secrecy or sexually explicit conduct. The prohibition is subject to statutory exceptions, including certain narrowly confined customer-service or research tools and internal employee-productivity systems.

The law prohibits providing listed features to covered users unless the operator determines through permitted age-assurance methods that the user is not a covered minor. A covered minor is under 18 and known by the operator to be a minor. In the absence of applicable rules, the statute requires reasonable methods, more than one option including one not dependent on government identification or allowing anonymity, and says self-declaration alone counts only when the user declares they are a minor.

Age-assurance data may be used only for that determination and must be deleted immediately after an attempt, subject to legal exceptions. The attorney general may seek injunctions, restitution, damages and civil penalties of up to $25,000 per violation. The law applies to covered conduct in New York; its text and future regulations govern the precise obligations.

Source: New York Governor ↗ · Source: New York State Senate ↗ · Source: New York State Senate ↗

Why does it matter?

Operators serving New York users need to assess both conversational behavior and age-assurance flows before the effective date. Families should not infer that every chatbot is prohibited, or that the new feature restrictions are already in force.