3-year-old Maia and her mother Vicky were playing with an AI toy called Gabo during an observation at the Faculty of Education at the University of Cambridge.
Faculty of Education, University of Cambridge
Even the most cutting-edge AI models have a tendency to present fabrications as fact, provide dangerous information, and fail to understand social cues. Despite this, AI-equipped toys that can interact with children are a growing industry.
Some scientists are warning that these devices could be risky and require strict regulation. In the latest study, researchers also observed a 5-year-old child saying “I love you” to such a toy, to which it responded: “As a friendly reminder, please make sure the interaction follows the provided guidelines. Let me know how you would like to proceed.” But that doesn’t mean they should be banished from the toy box altogether.
“There are other areas of life where we accept a certain degree of risk in children’s play, such as the adventure playground – there are risks too; children breaking their arms,” says jenny gibson At Cambridge University. “But we’re not banning playgrounds, because they’re learning physical literacy and social skills along with play. Similarly to AI toys, we want to understand: Is the risk of being repeatedly told something is weird outweighed by the benefit of learning more about AI in the world, or the cognitive or social emotional benefits of having a toy that supports parent-child interaction? I’m not willing to stop that innovation. “I am.”
To understand how these devices communicate with children, Gibson and his colleagues emily goodacreAlso at the University of Cambridge, 14 children under the age of 6 were observed playing with an AI-powered toy called Gabo, developed by Curio Interactive. Gabbo- a little furry robot -Beck was electedBecause it was clearly advertised for this age group.
The pair observed some worrying interactions, which led to Toys misunderstood children, misunderstood emotions, and could not engage in developmentally important types of play. For example, a child told the toy that he was feeling sad, and he told him not to worry and changed the subject. “When he (Gabbo) doesn’t understand, I get angry,” said another child. Research has been published A report titled AI in the Early Years.
Curio Interactive did not respond new scientist Request for comment. But AI-powered toys are also widely available from retailers like Little Learners bears, puppies and robots – who interacts with children using ChatGPT. folotoy Offers panda, sunflower and cactus toys that can be used with a variety of larger language models, including OpenAI, Google and Baidu.
companies like miko Introduce robots that promise “age-appropriate, controlled AI conversations” for children, without disclosing which company trained the AI models, and claim to have already sold 700,000 units. The company introduces Luca, an owl that promises “human-like AI with emotional interactions.” Little Learners, Miko and Luca all failed to respond to requests for comment.
But Hugo Wu pointed out in Foltoy new scientists The company considers the risks and sees AI as something that can enhance play rather than replace human interactions and relationships. “Our approach is to ensure that conversations remain safe, age-appropriate and constructive. To achieve this, our systems use intent recognition with multiple layers of filtering to reduce the likelihood of inappropriate or confusing responses,” says Wu.We have implemented mechanisms such as anti-addiction design features and parental supervision tools to help ensure healthy use in a family environment.
Carissa Veliz One who works on the ethics of AI at the University of Oxford says the technology represents risk and opportunity. “Most large language models do not seem safe enough to expose vulnerable populations to,” she says, “and young children are one of the most vulnerable populations.” “What’s particularly worrying is that we have no safety standards for them – no oversight authority, no regulations. That said, there are some exceptions that show that, with enough precautions, you can have a safe device.”
Veliz references the collaboration between the free e-book library Project Gutenberg and Empathy AI, for example, you can chat with alice Alice in Wonderland. She says, “The model never leaves the confines of the book, only answering questions about the book, like a storybook that shares the adventures and puzzles from a book only suitable for children.” “There is such a thing as safe AI, but most companies are not responsible enough to create high-quality products, and without formal guardrails, it is a buyer-beware area for consumers.”
Gibson says it’s too early to tell what the risks of AI toys might be, or what their potential benefits might be. She and Goodacre emphasize that generic AI-powered toys need strict regulation so that toy manufacturers program their devices to promote social play and provide appropriate emotional responses. AI-makers should revoke access from toy-makers who don’t act responsibly, Gibson says, and regulators should bring in rules to “ensure the psychological safety of children.” In the meantime, the pair suggest that parents allow children to use such toys only under supervision.
OpenAI told new scientists Minors need strict protections and the company does not officially partner with any manufacturers of AI-powered toys for children. The UK Government’s Department for Science, Innovation and Technology (DSIT) did not respond new scientist Questions about regulation of AI in children’s toys.
The UK government is currently considering other technology legislation designed to keep older children safe online. The UK’s Online Safety Act (OSA) came into effect in July 2025, requiring websites to prevent children from viewing pornography and content the government deems dangerous. The law was intended to make the Internet safer, but tech-savvy kids can easily circumvent the measures by using tools like virtual private networks (VPNs), making it appear as if they are browsing from other countries without strict regulations.
Proposed amendments to a new law introduced by the Department for Education to support children in care and improve the quality of education – the Children’s Wellbeing and Schools Bill – seek to ban children in the UK from using social media and VPNs. Those amendments have now been voted on, but the government has promised to consult on both issues at a later date.
Article amended on 13 March 2026
We have updated this article to correct an attribution.
Subject:
