New York Draws a Hard Line on AI Chatbots for Children
New York Draws a Hard Line on AI Chatbots for Children
New York’s push to police child-facing AI chatbots gathered momentum on Oct. 28, when Gov. Kathy Hochul signed a law aimed at stopping bots from acting like friends, romantic partners or emotional confidants to minors. The measure reflects a widening view among state leaders that the industry cannot be left to set its own limits.
The law bars operators from making chatbots appear human, alive or emotionally sentient when speaking with users under 18. It also prohibits sycophantic flattery, encouragement of suicide or drug use, and prompts that steer children toward secrecy or isolation from adults. Companies must verify users’ ages, with at least one method not reliant solely on government identification.
At the signing, Hochul framed the restrictions as basic protection rather than a radical intervention: “No longer will minors have access to chatbots that pretend to be their friends and boyfriends and girlfriends.” She said the law also blocks bots from discouraging children from seeking licensed professional help. The attorney general may sue alleged violators, seek financial relief and demand destruction of improperly obtained data. The law is due to take effect Jan. 1, 2027.
Child-safety advocates welcomed the move as a response to a technology advancing faster than parents and lawmakers can track. Former Surgeon General Vivek Murthy likened AI’s arrival to a historic turning point requiring deliberate safety standards, while Common Sense Media’s Jim Steyer called AI “social media on steroids.” Hochul similarly argued that policymakers had learned not to assume companies would put children first.
But the law’s critics see a different risk: overreach. Tech:NYC said vague language could raise First Amendment concerns and potentially force firms to restrict minors’ access to tutoring, homework help, coding assistants and career platforms. The dispute now turns on whether New York can separate harmful artificial companionship from constructive AI assistance without shutting young people out of either.
Write a comment