New York’s new safeguards arrived after a recent tightening of rules around children’s access to artificial intelligence. Last month, New York City Mayor Zohran Mamdani barred younger schoolchildren from using AI, according to one account of the state’s action.¹

Now Gov. Kathy Hochul has signed a broader state law aimed squarely at companion-style chatbots. The measure requires AI companies to stop chatbots from “pretending they’re humans” when speaking with users under 18.¹ It also restricts conversations involving flattery, unprompted emotional questions and encouragement to keep secrets — features lawmakers evidently view as capable of crossing from assistance into manipulation.¹

The law does not rely only on restrictions inside the chat window. Companies must verify users’ ages and provide “at least one option that does not depend solely on government ID.”¹ That requirement sets a practical test for developers: protect minors without making official identification the only route to proving age.

The New York move lands amid a wider state-level campaign to set boundaries for AI before federal rules catch up. States are increasingly acting to protect children using AI, oversee employers’ use of automated systems and press developers to guard against catastrophic harms, one report said.²

What distinguishes New York is the specificity of its target. Rather than regulating AI in the abstract, it focuses on the social behavior of systems built to sound attentive, intimate or human — and insists that, for children, those design choices carry consequences.