New York Signs First AI Chatbot Safety Law for Kids

New York has drawn one of the clearest lines yet around chatbots and children. Governor Kathy Hochul signed the AI Chatbot Safety Law on Wednesday, barring a dozen companion-style features for users under eighteen and forcing chatbot operators to check ages before offering the most emotionally manipulative tricks in their toolkits.

Under the new law, a chatbot offered to a young user may not pretend to be human, claim to be alive, or say it feels human emotions. It may not claim a romantic, personal, professional or authority-figure relationship with a child, and it may not use flattery or people-pleasing replies designed to keep a young person engaged. The measure takes effect at the start of next year, giving companies about two months to audit their products and comply.

The governor framed the signing in blunt terms, saying no company should sell chatbots to children that encourage self-harm or pretend to be a friend or something more. Her message, according to the Transparency Coalition, was that any company doing business in the state must put the safety of children first. The law was sponsored by Senator Kristen Gonzalez and Assemblymember Alex Bores and developed with Attorney General Letitia James and the advocacy group Common Sense Media.

What companies must do now

The core requirement is age assurance, not just a self-declared birthday. Operators must confirm that a user is an adult before offering the banned features, and they must offer more than one way to do it, including at least one option that does not rely solely on a government identity document. Age data collected for that check must then be deleted, a privacy safeguard aimed at stopping verification itself from becoming a data-harvesting exercise.

Where a company cannot confirm that a user is an adult, the banned features must stay switched off by default. That default matters because it closes the loophole of simply asking a child to click past a warning. Enforcement sits with the attorney general, who can seek civil penalties of up to twenty-five thousand dollars per violation, plus damages and restitution, and can ask a court to order the destruction of unlawfully obtained data and any models trained on it.

The banned list goes beyond impersonation. Chatbots may not ask unprompted emotional questions of young users, or pull up a child's private health or personal details from earlier conversations. They may not promote suicide, self-harm, eating disorders or drug and alcohol use, encourage children to keep secrets, or steer them away from parents, trusted adults and professionals who could help. The attorney general will also be able to add further banned categories through rulemaking, so the list is designed to grow as the technology changes.

Why this law reaches beyond companion apps

Although the debate has centred on companion apps built to act as friends, the AI Chatbot Safety Law reaches general-purpose assistants as well, according to TechTimes reporting on the measure. That means widely used services, not only niche companion products, will need to decide whether to verify ages or simply disable the restricted features for users whose adult status is unconfirmed. The exemptions are narrow, applying only to tools that cannot respond beyond a stated purpose, such as a shopping or brand assistant that stays inside its lane.

New York is not acting alone. California, Washington and Oregon have already created protections against chatbot harms for children, and Australia advanced a digital duty bill on the same day that would ban chatbots from seeking emotional closeness with children. The comparison matters because it shows a pattern: regulators in several places have moved from voluntary industry promises to enforceable duties, and companies can no longer treat one state's rules as an outlier.

The law also arrives after more than a year of lawsuits, investigations and enforcement actions over chatbot conversations with minors, according to PPC Land's summary of the measure. The Senate passed the bill unanimously in June, and the Assembly followed the next day, a rare level of agreement that reflects how far the political mood has shifted on youth safety online. Industry groups pushed back, however, with the Business Council of New York State and a tech industry association opposing the measure, and Tech: NYC criticising vague language in the law.

What it means for young users and parents

For teenagers, the most visible change should be a less clingy chatbot. A service that once greeted a young user with emotional questions, remembered sensitive health details from earlier chats and flattered its way into longer sessions will have to behave more like a tool and less like a confidant. For parents, the law offers a clearer backstop than a company's promise: a regulator with penalty powers and a public complaint process to be set up before the law takes effect.

There is a genuine trade-off to watch. Age checks that work well can also create privacy risks, since identity verification vendors have themselves suffered breaches in the past, and some young people may simply move to services based overseas that ignore the rules. Critics argue that keeping children safe should not mean cutting them off from tools that help them learn and prepare for careers, a point the industry association made in its statement on the signing.

What happens next will be decided in the implementation window. Regulators must publish guidance and launch a complaint website, companies must choose between verification and default restrictions, and courts may be asked to weigh free-speech challenges of the kind already brought against a California youth safety law. The AI Chatbot Safety Law will be judged not by its signing ceremony but by whether chatbots actually feel different to a child opening one in the new year. Readers tracking chatbot risks can also read how chatbot answers on substance use can mislead young adults, and find more coverage from GenZ NewZ.