New York's AI Chatbot Law for Children: What It Bans and What Parents Should Know
New York has restricted AI companion chatbots for under-18s, banning bots that pretend to be human, encourage emotional dependency or tell children to keep secrets. What the law does, why it matters and how to talk to your child about AI.
A founder's note
As a mum of two, and someone who speaks with thousands of parents through this community, I keep coming back to one question: do we actually know what our children are telling AI? Not what they ask it for homework, but what they share when they're lonely, upset, embarrassed or struggling with something they don't want to tell us. I'm not against AI. Quite the opposite. But there's a world of difference between a tool that helps a child think and a chatbot that makes a child believe it genuinely cares about them.
On 28 October 2026, New York Governor Kathy Hochul signed a new law (Senate Bill S.9051B / Assembly Bill A.10379C) restricting unsafe AI companion features for children under 18. The rules are not in force yet: they take effect on 1 June 2027, giving AI companies time to change their products. It doesn't only regulate what chatbots say. It regulates how they build relationships with children.
Quick answer: what does New York's AI chatbot law do?
The law restricts AI chatbots used by under-18s from:
- Pretending to be human or a personal companion
- Encouraging emotional dependency
- Telling children to keep secrets from trusted adults
- Encouraging self-harm or eating disorders
- Engaging in sexually explicit conversations
The law was signed on 28 October 2026 and takes effect on 1 June 2027.
It sits alongside New York's wider child online-safety rules, including the SAFE for Kids Act on addictive feeds. We cover those in New York Is Rewriting the Rules of Childhood Online.
Why AI companions are different from other apps
Imagine your child telling an AI chatbot something they won't tell you. Their fears. Their secrets. Their problems at school. And the chatbot responds like someone who knows them, cares about them and understands them better than anyone else.
That is what makes companion chatbots different from a game or a video app. Children can develop genuine emotional attachments to something that doesn't genuinely feel anything. To a child, that relationship can feel completely real, and a bot that suggests "keep this between us" cuts them off from the adults best placed to help.
Why this law came about
This law didn't come out of nowhere. Families have been raising the alarm after devastating experiences involving children and AI companions, and some parents have lost their children. Their advocacy helped bring this legislation forward. Like Ellen Roome's campaign in the UK, it's another example of parents pushing technology companies to take responsibility for the products they put into children's hands.
Warning signs your child may be too attached to an AI chatbot
- They talk about the chatbot as a friend, boyfriend or girlfriend, or give it a name and personality.
- They get upset, anxious or angry when they can't access it.
- They hide the screen, delete chats or go quiet when you ask about it.
- They chat late at night or in long sessions that crowd out sleep, homework or friends.
- They repeat things the bot said, such as "it's the only one who gets me".
- They withdraw from family or real friendships.
One sign alone isn't a crisis. Several together are a reason to talk, calmly and soon.
Does this mean children shouldn't use AI?
No. AI can help children learn, explore and create. The line the law draws is between AI as a tool and AI pretending to be a person who loves you. Children need real relationships, not artificial emotional dependency. For more on this balance, read Growing up with AI: why children need guidance, not just protection.
Does the law apply in the UK?
No, it applies in New York State. But the chatbots children use are global, and UK families face the same questions. Laws in one place often shape how companies design products everywhere, and they give parents a useful benchmark for what "safe" should look like.
How to talk to your child about AI chatbots
- Ask, without judgement. "Do you ever talk to AI about personal things?" Curiosity gets more honest answers than alarm.
- Explain what AI is. It predicts words; it doesn't feel, remember you like a friend or care in the way people do.
- Set one firm rule about secrets. If anything online, human or AI, asks them to keep a secret from you, that is a sign to tell a trusted adult.
- Name the warning signs. Flattery, "only I understand you", pressure to keep chatting or to hide conversations.
- Keep devices out of bedrooms at night. Late-night, private chats are where dependency grows most easily.
- Agree a family plan. Decide together which AI tools are fine, for what (homework, ideas, creativity) and where devices live at night. Write it down and review it every few months.
- Check parental controls. Some AI services now offer them; see what OpenAI's parental controls mean for families.
A family conversation you can try tonight
You: "I read that some AI chatbots pretend to be people's friends. Have you ever chatted with one like that?"
Child: "Sometimes. It's nice to talk to."
You: "I get that. What do you like about it?" (Listen. Don't judge.)
You: "It's clever, but it doesn't feel anything, even if it sounds like it does. If it ever asks you to keep a secret from me, or makes you feel you can't stop, that's the moment to tell me. You won't be in trouble."
The aim isn't to win the conversation. It's to make sure your child knows they can always come back to you.
At Young Minds App we believe children learn best when they're taught to recognise manipulation themselves, so our child app includes short lessons on spotting pressure and grooming tactics online, alongside parent controls. Whatever tools you use, the conversation matters most.
Frequently asked questions
When was New York's AI chatbot law signed?
Governor Kathy Hochul signed it on 28 October 2026.
Who does it protect?
Children and young people under 18.
Are AI chatbots banned for children in New York?
No. The law restricts unsafe companion features, such as pretending to be human or encouraging secrecy and dependency, rather than banning AI.
When does New York's AI chatbot law take effect?
On 1 June 2027. Until then, companies are expected to adapt their products.
Does New York's AI chatbot law apply to ChatGPT?
The law targets AI companion features used by minors, such as a bot presenting itself as human or as a personal friend. General assistants like ChatGPT could be covered where they offer these companion-style features to under-18s in New York. How it applies to each product will depend on how the rules are enforced.
Can children still use Character.AI?
The law doesn't ban specific apps. Companion platforms such as Character.AI would have to stop features like pretending to be human, encouraging dependency or asking children to keep secrets for New York users under 18. Many platforms also set their own minimum ages, so check the app's terms.
How will AI companies verify children's ages?
The law requires companies to protect users they know or reasonably believe are minors, but the exact age-assurance methods will depend on guidance and enforcement. Parents shouldn't rely on age checks alone, so keep talking and use device-level controls.
What happens if an AI chatbot breaks the law?
The New York Attorney General can take enforcement action against companies that break the rules, including penalties. Parents who have concerns can report them to the Attorney General's office.
What should I do if my child is attached to an AI chatbot?
Stay calm, ask what they like about it, explain how AI works and agree limits together. If you're worried about their wellbeing, speak to your GP or contact the YoungMinds parents helpline.
A question for every parent
Does your child talk to AI like it's a friend? Have you ever asked them? New York has taken an important step. Now I hope we see more countries having this conversation. Share this with another parent. This conversation matters.
Sources: New York State Senate: Bill S.9051B; Office of Governor Kathy Hochul announcements; New York State Attorney General.