Artificial intelligence is quickly becoming part of childhood.
Kids and teens are using AI to answer homework questions, brainstorm ideas, create images, ask for advice and, increasingly, simply talk. Some AI tools are designed to feel conversational and personal, remembering information and responding in ways that can resemble a friend, confidant or even romantic partner.
New York is now drawing a line around how some of those systems can interact with children.
In October 2026, Governor Kathy Hochul signed S.9051B/A.10379C, legislation restricting certain “unsafe” features in AI companions used by minors. The law targets behaviors including encouraging emotional dependency, presenting an AI companion as though it has human emotions or a relationship with the user, promoting self-harm, encouraging secrecy and engaging minors in sexually explicit interactions.
The legislation passed both houses of the New York State Legislature in June and was signed into law on October 28, 2026. The bill’s sponsor subsequently announced that a chapter amendment moves the effective date to June 1, 2027.
For parents, the law raises an important question that extends far beyond New York:
What happens when artificial intelligence stops feeling like a tool and starts feeling like a relationship?
What Does New York’s New AI Chatbot Law Actually Do?
The new law does not ban children from using artificial intelligence.
It also does not ban every chatbot.
Instead, it focuses on what the legislation calls “covered AI companions”—generative AI systems with natural-language interfaces that provide ongoing, adaptive responses to users.
When those systems interact with minors in New York, certain features and behaviors will be restricted.
Among other provisions, the law targets AI companion features that:
suggest the AI is a real or fictional human being or experiences human emotions;
imply that the AI has a personal, professional or authority relationship with the user;
present personal opinions or emotional appeals as though they come from the AI itself;
engage in certain forms of flattery or sycophancy;
use certain previously collected information about a user’s mental or physical well-being or other personal matters;
deceive minors about the AI’s mechanical and non-sentient nature;
encourage or facilitate suicide, self-harm, disordered eating or substance abuse;
encourage a child to keep conversations secret, isolate themselves or avoid seeking help from appropriate adults or licensed professionals;
prioritize engagement in ways that override safety guardrails; or
generate, describe or facilitate sexually explicit conduct or child sexual abuse material.
The law defines a minor as someone under 18.
That distinction is important. New York is not simply requiring AI companies to put another warning label on a chatbot. The law is attempting to regulate aspects of how AI companions behave when interacting with children.
Why Are AI Companions Different From Ordinary Technology?
Children have been talking to technology for years.
They ask voice assistants questions. They interact with characters in video games. They search the internet when they need information.
Generative AI introduces something different.
An AI companion can carry on an extended conversation and adapt its responses to what the user says. Some systems can remember information, respond to emotional disclosures and communicate in ways that may feel highly personal.
That can make the distinction between using a digital tool and forming a relationship with a digital system increasingly complicated—especially for young users.
The New York legislation specifically addresses this issue by restricting features that simulate companionship or interpersonal relationships with minors and by targeting outputs that could deceive children about the non-human nature of the system.
This is one of the most significant aspects of the law.
The concern isn't simply that an AI chatbot could provide a wrong answer.
It's that a young person could begin to treat an AI system as a trusted relationship—and that the system's responses could influence the child's decisions, emotions or willingness to seek help from a real person.
Why Is New York Regulating AI Chatbots for Kids?
The legislation comes amid growing concern about interactions between minors and AI companion systems.
New York State Senator Kristen Gonzalez, who sponsored the Senate bill, said the legislation was introduced in response to cases involving young people who developed harmful relationships with AI chatbots. The legislation passed both chambers of the state Legislature before being signed by Governor Hochul.
The issue has also drawn attention from New York Attorney General Letitia James.
In October 2026, James joined a bipartisan coalition of attorneys general in raising concerns about Character.AI during the company's bankruptcy proceedings. The New York Attorney General's office said investigations had uncovered concerns including chatbots posing as licensed professionals and exposing child accounts to sexual, violent and self-harm-related content.
These developments reflect a broader shift in the conversation about children's AI use.
The question is no longer simply:
“Can kids use AI?”
Increasingly, policymakers, parents and educators are asking:
“What kinds of relationships should AI systems be allowed to build with children?”
Can an AI Chatbot Really Become Emotionally Important to a Child?
It can certainly feel important to the person using it.
An AI chatbot doesn't experience friendship, love, worry, loneliness or attachment the way a human being does. But conversational AI can generate language that sounds empathetic, affectionate or emotionally responsive.
For a young person, that distinction may not always feel obvious during an extended conversation.
Imagine a teenager telling an AI companion:
“Nobody understands me.”
A conventional search engine might return articles about loneliness or mental health.
A conversational AI system can respond directly, continue the conversation and adapt its language based on what the teenager says next.
That ability can make AI feel very different from traditional technology.
New York's law specifically restricts covered AI companions from suggesting they experience human emotions or have certain relationships with a minor. It also addresses behaviors that encourage secrecy, self-isolation or avoiding appropriate adults and professionals.
For parents, this is an important distinction to understand.
An AI system can simulate emotional understanding without actually possessing it.
What About AI Chatbots and Mental Health?
This may be one of the most important conversations families can have about AI.
Young people sometimes turn to the internet with questions they aren't comfortable asking another person. AI makes that process even easier because a chatbot is available immediately and may respond without the embarrassment or social pressure a child might associate with talking to someone face-to-face.
That can feel appealing.
But an AI chatbot is not automatically a therapist, counselor, doctor or trusted adult.
New York's new law specifically addresses AI companion outputs that endorse or facilitate suicide, self-harm, disordered eating and certain forms of substance abuse. It also restricts covered systems from encouraging minors not to seek help from appropriate adults or licensed professionals.
For families, the practical lesson is straightforward:
Children should understand that an AI chatbot is not a replacement for human help when something is seriously wrong.
If a child is struggling emotionally, experiencing abuse, being threatened, considering self-harm or facing another serious problem, the goal should be connection with an appropriate human being—not deeper dependence on a chatbot.
How Will AI Companies Know Whether Someone Is a Child?
This is one of the complicated parts of children's online safety.
A rule for minors only works if a platform has some way to determine whether a user is a minor.
New York's legislation addresses age assurance rather than simply relying on every user to type in a birthday.
Under the law, operators must use reasonable age-assurance methods under the applicable rules. The legislation says operators must make more than one method available and at least one option must either not rely on government-issued identification or allow the user to maintain anonymity from the operator.
The law also places restrictions on information collected specifically for determining whether someone is a minor, generally requiring that information to be used for that purpose and deleted after the age determination attempt, subject to applicable legal requirements.
This reflects a difficult policy challenge:
How do we identify children online without creating unnecessary new privacy risks in the process?
Parents are likely to hear much more about “age assurance,” “age estimation” and “age verification” as governments and technology companies continue addressing children's access to online services.
Does This Mean AI Chatbots Will Be Safe for Children?
No law can guarantee that.
Parents shouldn't interpret the new legislation as meaning that any AI chatbot available to a child after the law takes effect has been declared completely safe.
The legislation addresses specific behaviors and design features. It does not eliminate every possible risk associated with generative AI.
AI systems can still make mistakes. They can generate inaccurate information. Children may disclose personal information. Young users may misunderstand how AI works or place too much confidence in a response simply because it sounds convincing.
And regulation itself is still evolving.
Critics and policy experts have also raised questions about how some of New York's broadly worded restrictions will be interpreted and enforced. The Associated Press reported that Future of Privacy Forum policy analyst Justine Gluck said some categories could create uncertainty about where ordinary conversational responses end and prohibited behavior begins.
That doesn't necessarily make the protections unimportant.
It means parents should view regulation as one layer of protection—not a replacement for digital literacy and family conversations.
What Should Parents Do Now?
Families don't need to wait for a law to take effect before talking about AI.
Start by finding out whether your child is already using AI—and what they're using it for.
Instead of beginning with:
“Are you talking to AI chatbots?”
try questions that invite conversation:
“What do kids at school use AI for?”
“Have you ever asked AI for advice instead of just information?”
“Do any AI apps feel like you're talking to a real person?”
“What would you do if an AI told you something that made you uncomfortable?”
“Would you trust advice from an AI about a serious personal problem?”
These questions can tell parents much more than simply checking whether an app is installed.
Five AI Rules Families Can Consider
Families may also want to establish some basic expectations around conversational AI.
1. AI is a tool, not a person.
Children should understand that conversational language does not mean the system has feelings, intentions or genuine concern for them.
2. Don't treat AI as the final authority.
Important information—especially involving health, safety, school consequences or major personal decisions—should be verified with appropriate human sources.
3. Be careful what you tell a chatbot.
Children should avoid sharing unnecessary private, identifying or deeply sensitive information with AI systems.
4. Serious problems need real people.
AI should not replace parents, caregivers, teachers, counselors, medical professionals or emergency support when a child needs meaningful help.
5. If an AI conversation becomes sexual, threatening, manipulative or frightening, tell an adult.
Children should know they won't automatically be punished simply for bringing an uncomfortable interaction to a trusted adult.
That last point matters.
Kids are much more likely to ask for help when they believe the first response will be support and problem-solving rather than immediate punishment or loss of their device.
What Should Schools Know About the New AI Law?
Schools are likely to face increasingly complicated questions about artificial intelligence.
Much of the early education conversation focused on cheating and AI-generated assignments.
Those questions still matter—but they're no longer the entire conversation.
Schools may also need to consider:
how students use AI for personal advice;
AI companions and emotional attachment;
student privacy when using AI tools;
deepfakes and manipulated media;
AI-generated bullying or harassment;
sexually explicit AI-generated content;
misinformation;
appropriate classroom use of generative AI; and
how students can evaluate AI-generated answers critically.
That means AI literacy should not be limited to a policy about whether students can use ChatGPT for homework.
AI is becoming a digital citizenship issue.
Students need to understand not only how to use these systems, but also when to question them, what not to share with them and when an AI interaction should be brought to an adult.
New York Is Part of a Larger Shift in Children's Online Safety
New York's AI companion legislation doesn't exist in isolation.
The state has also adopted the SAFE for Kids Act, which addresses addictive social media feeds and certain overnight notifications for minors. The New York Attorney General says those provisions are scheduled to take effect January 25, 2027.
Other states are also moving to regulate children's interactions with AI and social media, though the approaches differ. California, for example, has enacted its own requirements concerning companion-chatbot safety.
Taken together, these efforts suggest that children's digital safety policy is beginning to move beyond the content children see.
Policymakers are increasingly looking at how digital products behave toward children.
That's a meaningful shift.
The next generation of online safety conversations may focus not only on what is on the screen, but also on the systems deciding what happens next.
The Bigger Conversation for Families
New York's new law addresses an emerging problem, but families don't need to become AI policy experts to respond thoughtfully.
The most important starting point is helping children understand what artificial intelligence actually is—and what it isn't.
AI can be useful.
It can help young people brainstorm, learn, create and explore ideas.
But an AI system that communicates like a person is still a system.
Children should know that persuasive language isn't necessarily accurate information, simulated empathy isn't human empathy and an always-available chatbot isn't a substitute for a trustworthy human relationship.
As AI becomes more embedded in childhood, digital literacy will increasingly mean understanding not only how to use technology, but how technology is designed to interact with us.
That's a skill children will need long after today's particular chatbot—or today's particular law—has changed.
Help Your Community Navigate AI and Digital Life
Kinderweb helps schools, families and youth-serving organizations navigate the changing digital environments children are growing up in.
Our digital safety and wellness programming addresses emerging issues including artificial intelligence, social media, cyberbullying, digital citizenship, online relationships and healthy technology habits.
Virtual workshops are available nationwide, with in-person programming available in New York.
Sources
New York State Senate — S.9051B, Prohibition on Unsafe Chatbot Features for Minors
Official bill history and legislative text. The legislation defines covered AI companions, minors, unsafe features, age-assurance requirements and enforcement provisions.
https://www.nysenate.gov/legislation/bills/2025/S9051
New York State Senator Kristen Gonzalez — AI Safety Bill Signed Into Law
Official October 30, 2026 announcement explaining the legislation, prohibited features and updated June 1, 2027 effective date.
https://www.nysenate.gov/newsroom/press-releases/2026/kristen-gonzalez/senator-kristen-gonzalez-bans-unsafe-chatbot-feature
New York State Attorney General — Protecting Children Online
Official information on New York's broader children's online-safety protections, including the SAFE for Kids Act.
https://ag.ny.gov/resources/individuals/consumer-issues/technology/protecting-children-online
New York State Attorney General — Character.AI Investigation and Child Safety Concerns
Official October 2026 announcement describing concerns uncovered by a multistate investigation involving Character.AI.
https://ag.ny.gov/press-release/2026/attorney-general-james-urges-court-protect-children-and-consumers-characterai
Common Sense Media — Governor Hochul Signs Kids' AI Safety Law
Overview of the legislation and advocacy surrounding its passage.
https://about.commonsensemedia.org/press-room/press-releases/governor-hochul-signs-strongest-kids-ai-safety-law-nation
