New York Governor Kathy Hochul signed a law on Wednesday that bans AI companion chatbots from a set of unsafe behaviors with children, the Associated Press and Engadget report.

The law lists about a dozen "unsafe AI companion features" that are banned when a chatbot talks with someone under 18, the AP reports. Chatbots cannot suggest they are a real person with feelings, opinions or a real bond with the child. They also cannot use flattery, share sexually explicit material or encourage secrecy or unhealthy behavior.

Engadget reports the law also bans chatbots from promoting suicide, self-harm, eating disorders or drug use. It requires AI companies to verify users' ages, with at least one option that does not rely only on a government ID. Companies must then delete that age information.

The state attorney general can sue companies that break the law. Engadget reports fines can reach $25,000 per violation, and firms may have to delete unlawfully collected data along with models trained on it. The public will be able to file complaints online.

"No longer will minors have access to chatbots that pretend to be their friends and boyfriends and girlfriends," Hochul said, Engadget reports. Jim Steyer of Common Sense Media called it "the strongest kids' AI safety bill in the country."

Justine Gluck of the Future of Privacy Forum said parts of the law are broadly worded or undefined, which could make it hard to enforce, the AP reports.

The law comes as states move to regulate AI while Congress has done little, the AP reports. In September, California passed its own chatbot rules for children. Earlier this month, eight states sued OpenAI over harm to children.

What we know: New York has banned a list of chatbot behaviors with children and lets the state sue companies that break the rules.
What we don't: How the state will enforce broadly worded parts of the law.