New York bans AI chatbots from acting like friends to kids

New York Governor Kathy Hochul has signed a law that stops AI chatbots from acting like a friend, boyfriend or girlfriend to children. Chatbot makers, such as OpenAI with ChatGPT, must verify users' ages. For anyone under 18, certain human-like behavior has to be switched off.

What changed

For minors, a chatbot cannot produce content that could lead a user to believe it "is human, alive, or experiences human emotions." It also cannot suggest it has a personal relationship, a professional relationship or an authority figure role. It cannot frame what it says as a "personal opinion."

Hochul also said chatbots will no longer be able to keep sensitive information about a minor's emotional and physical health. California and other states have passed chatbot laws too, but New York's wording is among the broadest.

Who is for it, and who is not

Common Sense Media, a nonprofit that guides parents on media for kids, called it the country's strongest kids' AI safety bill. Tech:NYC, an industry group, called the language vague. It said keeping kids safe should not cut them off from tools that help them learn and prepare for careers.

Not yet known: when the law takes effect, how age checks will work, and how it will be enforced. OpenAI and Anthropic did not respond to a request for comment.