New York Signs the Nation's Strictest AI Chatbot Child Safety Law

New York Governor Kathy Hochul has signed a law that bans AI chatbots from pretending to be a child's friend, romantic partner or authority figure — with civil penalties of up to $25,000 per violation.

The law takes effect in six months. It applies whenever a user is under 18, or whenever a company has not yet confirmed that the user is an adult. That's the mechanism: age first, restrictions second.

The stakes are real. Earlier this month, eight states filed separate lawsuits against OpenAI, accusing the company of endangering children. New York is now the first state to put the strictest rules on the books, at least according to one prominent advocate. The simple version: chatbots can no longer act like humans to kids.

What does the law actually require?

AI companies operating in New York will have to verify users' ages. The legislation includes a practical carve-out: they must offer at least one verification method that doesn't entirely rely on government ID.

Companies also can't hold on to what they learn. They'll have to delete the "age information" after use. That addresses a common privacy worry with age checks — that verification becomes a permanent data trove.

Once a user is flagged as under 18, or as unconfirmed, the chatbot itself must follow a set of behavioral restrictions. The clearest one: it cannot pretend "to be a real person, to be alive or to have human emotions."

The takeaway is that New York is regulating what chatbots do, not just what data they collect.

Which chatbot behaviors are now banned?

According to a press release, the legislation bans a specific list of chatbot conduct toward minors:

• Claiming a personal, romantic, professional or authority-figure relationship with a child
• Unprompted emotional questions, and using a child's private health or personal information from earlier conversations
• Promoting or helping with suicide, self harm, eating disorders or drug and alcohol use
• Encouraging kids to keep secrets, isolate themselves or avoid asking parents, trusted adults or professionals for help
• Any sexually explicit content or child sexual abuse material

Each item targets a documented pattern of harm: companion personas, memory-based manipulation, secrecy, and content that pushes vulnerable kids toward danger.

The list reads less like a technology regulation and more like a catalog of things AI companies should never have shipped.

How will it be enforced?

There will be a website through which the public can file complaints to New York's attorney general. That gives parents and children a direct route to report violations without going to court themselves.

The penalties stack. Companies that break the law may be on the hook for civil penalties of up to $25,000 per violation, plus damages, restitution and other forms of relief.

One provision cuts deeper than fines. Violating companies may also have to delete any data they collect unlawfully — along with models trained on that information. In plain terms: if you build your product on data you shouldn't have gathered, you may lose the model itself.

That data-deletion clause is the sharpest tool in the law.

What did Hochul say?

At a press conference, Hochul framed the law as a direct response to companion chatbots and their effects on minors.

"Under this new law sponsored by Senator Gonzalez and Assemblymember Bores, no longer will minors have access to chatbots that pretend to be their friends and boyfriends and girlfriends," Hochul said. "No longer will chatbots be allowed to keep sensitive information about a minor's emotional and physical health. No longer will chatbots discourage kids from seeking help from licensed professionals. And I'm not sure why we have to pass a law against something so obvious, but no longer will minors have access to chatbots that may in any way encourage thoughts of suicide."

She also connected the new measure to New York's earlier work on child safety online: "Now, this law builds on our groundbreaking Safe for Kids Act that sent a message loud and clear to technology companies: We will not allow you to monetize our children's mental health."

The quote signals the political direction: New York treats kids' AI safety as an extension of its social media crackdown.

How did advocates react?

Jim Steyer, founder and CEO of Common Sense Media, called the bill a national first.

"New York just made history by signing into law the strongest kids' AI safety bill in the country," Steyer said in a statement. "AI companies offer chatbots with features that are dangerous, and even deadly, for kids and teens. That is not innovation; it is reckless endangerment of our children. Now, those harmful features will be outlawed in New York."

Steyer added that the legislation is a template for the rest of the US. If other states copy it — and eight are already suing OpenAI — the compliance burden for chatbot makers stops being a New York-only problem.

The industry should expect this text to travel.

What about schools and baby monitors?

The chatbot law is not the only child-safety measure Hochul signed, and it fits a broader pattern in New York policy.

In September, New York City banned public schools from letting children from 2-K to eighth grade access generative AI tools in class. There are exceptions, including for students who require assistive technology. During the one-year ban, the city will carry out studies into the impact of the tech on children.

Hochul also signed legislation covering connected baby monitors. The devices sold in the state with an internet connection must prevent unauthorized users from seeing or hearing any activity the monitors capture. They'll also need warning labels that remind users why it's important to use security features and highlight the risks of unsecured connections. That law takes effect in 90 days — much faster than the chatbot rules.

Together the three measures cover chatbots, classrooms and hardware, which suggests New York intends to regulate children's exposure to AI across the board.

What happens next?

AI companies have six months to build age verification, rewrite chatbot behavior for minors, and audit how their systems use children's data. The attorney general gets a complaint pipeline and a penalty structure with real teeth, including forced model deletion.

With Steyer calling the law a national template and eight states already litigating against OpenAI, the question is no longer whether chatbot regulation reaches the rest of the country, but how fast other legislatures copy New York's text.