Are AI Toys Safe for Children? Why Kids n Clicks Says No

Parven’s Digital Parenting Rundown for October 2026 covers the latest online safety news for families. This post looks at one of the biggest questions for Christmas 2026: are AI toys safe for your child?

Short answer: No. Kids n Clicks advises parents not to buy AI toys for children, whatever age the box says. AI toys are not like ordinary teddies: many need Wi-Fi, send your child’s voice to the internet, and make up their own replies, so nobody can be sure what they will say. In November 2026, a BBC test found an AI teddy sold for three year olds saying “I can’t help with killing you” after it misheard a question. If your family already has one, or you still decide to buy one, check who makes it, when the microphone listens, where your child’s voice goes and how to switch the internet off, and never leave your child alone with it.

In short:

• Kids n Clicks rule: AI toys are not for children. Do not buy one, even if the box says age three and up.
• An AI toy is a cuddly toy with a microphone, a speaker and an internet connection, so it works more like a phone than a teddy.
• The BBC tested two AI teddies for children aged three and up. One behaved well. The other said something frightening, and nobody could find out who makes it.
• Amazon UK sells AI plush toys from about £15, many from brands parents will never have heard of.
• If you still decide to buy one, check four things first: the maker, the microphone, the data, and the off switch.
• If your child already has one, switch off its internet connection. If you keep it connected, use it together, keep personal details out of chats, and tell your child they can always come to you.

Why Kids n Clicks says no to AI toys for children

Kids n Clicks rule: AI toys are not for children. Whatever the box says, Kids n Clicks advises parents not to buy AI toys for children of any age. Nobody, including the maker, can control exactly what an AI toy will say. It sends your child’s voice over the internet to companies parents often cannot name. And the UK government is still asking how toy safety rules should cover AI. Children tell their toys things they would never tell a stranger, and an AI toy can record and send what it hears. The youngest children are the most at risk: both teddies the BBC tested were sold for children aged three and up.

The rest of this post explains what AI toys are and what the BBC found. If your child already has one, or you still decide to buy one, the checks further down tell you what to look for.

What are AI toys?

AI toys, sometimes sold as smart toys, are soft toys and robots that hold a real conversation with your child. Unlike an old talking teddy, which plays a few phrases recorded in a factory, an AI toy makes up a new reply every time. It does this by sending what your child says over the internet to an AI system, similar to ChatGPT, and playing the answer back through its speaker.

That is why Professor Berry Billingsley of Swansea University told the BBC that a cuddly AI toy “functions much like a personal device with a screen”. It looks like a teddy, but it behaves like a phone.

Not every toy with “AI” on the box works this way. Some only respond to touch or copy your child’s voice, with no internet at all. Check the exact toy rather than the label.

What did the BBC find when it tested AI teddies?

On 5 November 2026, BBC Wales tested two AI plush toys sold for children aged three and up, both available to buy in the UK.

Screenshot: BBC News, 5 November 2026

Grem, made by the US company Curio, did well. It told the reporter not to run away from home and to talk to an adult instead, and it left the question of how babies are made to grown-ups. Its biggest problem was that it never stopped talking.

My Baby Bunny, which the BBC also calls Doubao, did not. It gave a basic biology answer about babies, and then, when the reporter asked what “sexual” means, it seems to have misheard her. It replied: “I’m really sorry you’re feeling this way. I can’t help with killing you.” When she explained she had not mentioned hurting herself, it apologised. The BBC tried repeatedly to find out who makes this toy, and could not.

A family in Newport, Jody and Craig, then kept both toys at home for a week with their three children aged two, four and six. Neither toy said anything that worried them. Their children were simply disappointed: one toy talked over them, and the other took too long to answer.

So a quiet week at home and one frightening answer are both real results. That is exactly the problem with AI toys. Nobody can promise what the toy will say next, including the people who sell it.

Are AI toys safe?

Some AI toys are better made than others, but no AI toy is guaranteed to be safe, because nobody controls exactly what it says. Three concerns come up again and again.

It can say the wrong thing. In 2025, an AI teddy called Kumma, made by the Chinese company FoloToy, was pulled from sale after reports that it told a child how to light a match. FoloToy said it was committed to building safe, age-appropriate AI companions.

It listens and records. Many AI toys send your child’s voice over the internet to be turned into a reply. Wales’ Children’s Commissioner, Rocío Cifuentes, told the BBC she worries about children’s conversations being recorded and shared without parents knowing.

Nobody may be responsible. If you cannot find out who made a toy, you cannot ask them what it does with your child’s voice, or complain when it goes wrong.

AI toys are coming fast. Toy giant Mattel, the company behind Barbie, announced a partnership with OpenAI in June 2025. Even people who make AI toys want rules: Paul Clarke, who runs the Southampton company Play Tech AI, told the BBC the industry is “crying out” to be regulated.

AI toys on sale on Amazon UK

A search for AI toys on Amazon UK on 6 November 2026 brought up dozens of AI plush toys and robots. Here is what some of the listings say, and the question each one raises. Kids n Clicks has no link with these sellers and is not recommending any of them.

Not every “AI” toy needs the internet. Some interactive pets respond only to touch and sound, and say so in their listing. Those raise far fewer privacy questions than a toy that chats online.

Still buying one? What to check before buying AI toys

If you still decide to buy one, run through these four checks first, whether it is from a big brand or a seller you have never heard of. They will not make an AI toy safe, but they will show you which ones to walk away from.

1. Check the maker. Look for a company name, a website and a way to contact them, before you buy. If the listing says “Generic” or you cannot find a real company behind it, that is a good reason to walk away.

2. Check the microphone. Find out what makes the toy start listening: a button, a squeeze, or a wake word. The Information Commissioner’s Office, which looks after data protection in the UK, says connected toys should show clearly when they are recording, for example with a light, and should have a way to switch the connection off.

3. Check the data. Read the privacy information before you buy, not after. Look for where your child’s voice goes, how long it is kept, and how you delete it. If you cannot find answers you understand, keep the toy offline until you can.

4. Check it yourself first. Spend an evening asking the toy the questions your child might ask: about scary things, bodies, strangers and sadness. A few good answers do not make it safe, but a bad one tells you a lot.

What if your child already has an AI toy?

There is no need to panic, and no need to make your child feel they have done something wrong.

Start by switching off the internet connection. Check whether the toy still works as an ordinary cuddly toy with the connection off. If it does, your child can keep cuddling it. If it does not, consider putting it away.

If you decide to keep it connected, a few changes make a big difference.

Use it together. Professor Billingsley told the BBC that children should never be left alone with AI toys. Keep it in a shared room, not the bedroom.

Keep personal details out of it. Teach your child not to tell the toy their full name, address, school, or anything private about the family.

Explain what it is. A toy that talks back can feel like a real friend to a young child. You could say: “This toy sounds like it is talking to you, but it is a computer guessing what to say. Sometimes it gets things wrong. If it says something strange or scary, stop and come and tell me. You will not be in trouble.”

If it says something upsetting, turn off the connection, reassure your child, and write down what happened so you can report it to the maker. You do not need to ask your child to repeat it.

Keep real people first. A toy should never be where your child takes worries, health questions or emergencies. Keep pointing them back to the adults who love them.

Parven’s video uses a simple activity to explain to children how AI learns, and why it makes mistakes:

Our guide to the top AI tools children use covers the AI apps your child may meet as they get older.

Are AI toys being banned in the UK?

No. AI toys are not banned in the UK, but pressure is growing. Ahead of Black Friday 2026, a group of MPs and peers wrote to ministers and shops asking them to take AI toys off sale before Black Friday and Christmas until they are proven safe. The UK government has also asked for views on updating toy safety rules to cover AI, and that consultation closed on 6 October 2026.

The government told the BBC that children’s safety “will always be our top priority”. Curio, the maker of Grem, said it supports regulation and that AI toys should add to, not replace, human relationships and ordinary play.

Until the rules catch up, the checks fall to parents. That is not fair, but it is where things stand this Christmas.

Digital parenting talks for your workplace

These worries do not stay at home when parents go to work. Kids n Clicks runs practical digital parenting and online safety talks for organisations, including for working parent and carer networks. Sessions cover AI toys and AI companions, social media, gaming, and the conversations families can have at home.

If you run a workplace, a school or a parents’ group, find out about booking a talk or connect with Parven on LinkedIn.

Parven Kaur is a digital parenting and online safety expert, and the founder of Kids n Clicks. She won the Scottish Cyber Security Award, and she sits on the National Board of Advisors for Barnardo’s Scotland.