Susanna Riva New York Governor Kathy Hochul signed two bills this morning aimed at reshaping the way artificial intelligence enters young people's homes. The main measure concerns generative chatbots , which in recent months have become virtual companions for teens and tweens. The legislation is part of a broader state strategy aimed at limiting technologies potentially harmful to minors, adding to measures already passed in California and twelve other states.
Key provisions of the legislation on chatbots for minors
The law requires all AI-based chatbot providers to confirm the user's age before activating any features. Verification must be done using multiple methods, at least one of which cannot rely solely on a government-issued ID.
Once confirmed, age data must be deleted immediately, thus preventing the creation of longitudinal profiles of minors.
Prohibited content and ways to interact
The law lists twelve categories of behavior deemed dangerous. Bots cannot appear human , claim to have emotions, or present themselves as a reference figure—either personal or professional—to a user under 18.
The use of flattery, responses designed to keep the user glued to the screen, unsolicited emotional questions, or the disclosure of health or personal data collected in previous conversations is also prohibited. Furthermore, bots must not promote suicide, self-harm, eating disorders, substance abuse, or sexually explicit content, nor encourage minors to withhold information from parents or caregivers.
Sanctions and remedies provided
The state Department of Consumer Affairs can impose fines of up to $25.000 for each violation, as well as seek damages and order the destruction of illegally collected data. In cases of serious violations, authorities can also order the deletion of algorithms trained with minors' information, effectively limiting the responsible company's future use of those models.
Extension of the regulation to baby monitors
In addition to chatbots, the second bill focuses on internet-connected baby monitors . Starting ninety days after its publication, all new devices sold in the state will be required to include security protections that prevent unauthorized access to real-time audio or video. Additionally, each unit will be required to display a clear label warning of internet connectivity and prompting families to configure privacy settings. The measure was prompted by security tests that awarded eight out of nine devices an "F" or "D" rating, highlighting the vulnerability of these products to potential hackers.
Industry reactions and future prospects
The trade association Tech:NYC has denounced the overly broad language in the law, fearing potential conflicts with the First Amendment's right to freedom of expression. Meanwhile, AI companies are already updating their security policies, but concerns remain that broad definitions could also limit legitimate services, such as tutoring or coding assistance. In a national context where twelve states have passed similar regulations, New York is positioning itself as one of the most stringent, pushing for a "floor" of protection for children in the face of the rapid evolution of AI.