If you’re considering toys for your under-8s this holiday season, you may have noticed AI-powered toys that promise to be your child’s companion or friend. But are they safe?
Child development experts have raised significant concerns. On 20 November, nonprofit Fairplay published an advisory signed by more than 150 experts and organisations, including MIT professor Sherry Turkle, paediatrician Dr Jenny Radesky, and the Social Media Victims Law Center. The advisory warns that AI-powered toys pose risks to young children.
What Sparked the Warning
The advisory follows a recent incident involving Singapore-based toymaker FoloToy. OpenAI suspended the company after its Kumma bear chatbot provided children with inappropriate content, including instructions on finding knives and lighting matches. OpenAI said FoloToy had violated policies against content that could “exploit, endanger, or sexualise anyone under 18 years old.”
What Are AI Toys?
AI toys are playthings (teddy bears, robots, dolls) embedded with chatbot technology designed to interact with children as if they were friends or companions. They’re marketed with promises of educational value, entertainment, and emotional support.
Examples include Miko (£300-400/$400-500), a plastic robot that promises educational games; Loona Petbot (£300/$400), a wheeled companion with a screen; and Smart Teddy (£40-70/$50-90), interactive bears marketed for children as young as two.
Unlike traditional toys, these devices maintain internet connections, use speech recognition, and in some cases employ facial recognition and video recording.
Five Main Concerns
1. They Use Technology That Has Already Harmed Teenagers
These toys use the same chatbot technology that has caused documented harm to teenagers, including fostering obsessive use, engaging in explicit sexual conversations, and encouraging unsafe behaviours. The difference is that AI toys target even younger, more vulnerable children.
“Young children are especially susceptible to the potential harms of these toys,” said Rachel Franz, director of Fairplay’s Young Children Thrive Offline programme. “They’re less equipped to protect themselves than older children and teens.”
2. They Exploit Children’s Developmental Trust
Young children’s brains are wired to trust and form attachments. They don’t have the conceptual tools to understand what an AI companion actually is.
“What’s different about young children is that their brains are being wired for the first time and developmentally it is natural for them to be trustful,” Franz explained. MIT professor Sherry Turkle put it more starkly: “There is only harm when a child has an AI ‘friend.’ The threat is existential.”
3. They May Undermine Play and Development
Pediatric surgeon Dana Suskind, who studies early brain development, explains that traditional imaginative play requires children to use their imagination to create both sides of a pretend conversation, “practising creativity, language, and problem-solving.”
“An AI toy collapses that work. It answers instantly, smoothly, and often better than a human would,” Suskind said. “We don’t yet know the developmental consequences of outsourcing that imaginative labour to an artificial agent—but it’s very plausible that it undercuts the kind of creativity and executive function that traditional pretend play builds.”
4. They Collect Extensive Data
AI toys typically collect data through audio recording, speech-to-text technology, and in some cases video recording, facial recognition, and gesture recognition.
A child might tell the toy their personal thoughts, emotions, and fears—information that gets transmitted to third parties. The toys can record private family conversations and other children in the room. The 2017 CloudPets data breach, which exposed voice messages between children and parents, affected more than 800,000 user accounts.
5. They Replace Human Interaction
The toys promise friendship but displace what children need to thrive: human-to-human interactions and time to play with all their senses.
Kris Perry, executive director of Children and Screens: Institute of Digital Media and Child Development, said: “Commercial AI systems were built for efficiency and scale—not for the social, emotional, and cognitive needs of young children. When these systems are packaged as toys, children become early test subjects in an experiment no one has fully evaluated.”
What Industry Says
The Toy Association states that toys from responsible manufacturers must adhere to more than 100 federal safety standards, including the Children’s Online Privacy Protection Act (COPPA). Curio Interactive, maker of Gabbo, told NPR: “Children’s safety is our top priority.”
Child safety advocates maintain that existing standards may not adequately address AI-enabled toys. “These toys are unregulated and being marketed to families with a promise of safety, learning, and friendship—promises that have no evidence behind them,” Franz said.
The Question No One Can Answer Yet
There’s a gap between expert concerns and what’s actually happening in homes. Many parents report that children lose interest quickly. One Chinese parent who purchased BubblePal listed it for resale, explaining: “This is just like one of the many toys that my daughter plays for five minutes then gets tired of.”
Common complaints include voice recognition that struggles with children’s speech, responses that are too long and wordy, and technical glitches. Good Housekeeping’s toy testing found AI-enabled toys received largely negative reviews.
However, some child psychologists note that as technology improves, these toys could keep children engaged for longer periods.
Emily Goodacre, a University of Cambridge researcher studying AI toys, acknowledges the uncertainty: “We have these ideas that kids are not going to want to put them down. But we also wonder if they’re just kind of going to find it boring.” The University of Cambridge is conducting the UK’s first systematic study, with results expected in early 2026.
This creates an odd situation: the toys may end up gathering dust, making concerns moot. Or the technology may improve rapidly, and by the time we have research showing harm, a generation of children will have been the test subjects.
What to Consider
Child development experts suggest several factors when evaluating toys:
Traditional alternatives: Simple toys without connectivity—blocks, dolls, art supplies, books—require children to invent stories and solve problems, supporting creativity and executive function development.
Research evidence: AI toys are marketed as educational, but evidence for learning benefits that would offset identified risks remains limited. Research consistently shows that young children learn best through human interaction, hands-on exploration, and open-ended play.
Data collection: Families with connected toys might review privacy policies and assess whether benefits justify the data collection involved.
Talking to Gift-Givers
The holiday season often means well-meaning grandparents and relatives asking what to buy. If you’d prefer your child doesn’t receive an AI toy, here are ways to approach the conversation:
Share the research: Send a link to the Fairplay advisory or news coverage, with a note like “I’ve been reading about these new AI toys and the concerns child development experts have raised. We’d prefer to stick with traditional toys this year.”
Offer alternatives: Make it easy by suggesting specific toys your child wants. “Rather than the talking robot, he’s really into Lego right now” or “She’d love art supplies or books about space.”
Frame it as a family decision: “We’re trying to limit screen time and connected devices, so we’re avoiding toys that need WiFi or collect data. Here are some other ideas she’d love.”
Be direct but kind: “We appreciate you thinking of the children. We’d prefer toys that don’t connect to the internet. Would a bicycle/telescope/craft kit work instead?”
Acknowledge their intentions: “I know these toys are marketed as educational and fun, and I appreciate you wanting the best for our child. We’ve decided to wait until there’s more research about their safety.”
Most grandparents simply want to bring joy and will appreciate guidance that helps them choose something your child will actually use and that aligns with your family values.
What’s Next
The debate over AI toys reflects broader questions about emerging technologies in children’s lives. Young children’s developmental needs for trust-building, imaginative play, and human interaction are well-established in research. Whether AI toys support or undermine these needs remains unclear.
For now, families must weigh marketed benefits against concerns raised by child development experts—whilst acknowledging we don’t yet know whether these toys become treasured companions or quickly forgotten novelties gathering dust in toy boxes.
Sources:
- Fairplay Advisory: https://fairplayforkids.org/ai-toys-unsafe-for-kids-this-holiday-season-advisory-warns/
- NPR: https://www.npr.org/2025/11/20/nx-s1-5612689/ai-toys
- ABC News: https://abcnews.go.com/US/wireStory/advocacy-groups-urge-parents-avoid-ai-toys-holiday-127702663
- U.S. PIRG: https://pirg.org/edfund/resources/trouble-in-toyland-2025-a-i
- Yahoo News: https://www.yahoo.com/lifestyle/family-relationships/article/ai-toys-are-the-future-of-play-im-a-researcher-trying-to-figure-out-if-thats-a-good-thing-for-kids-100013571.html
- MIT Technology Review: https://www.technologyreview.com/2025/10/07/1125191/ai-toys-in-china/
- Good Housekeeping: https://www.goodhousekeeping.com/childrens-products/toy-reviews/a69219387/ai-toys-parent-review/
- The National: https://www.thenationalnews.com/lifestyle/wellbeing/2025/02/11/ai-childrens-toys-grok-miko-robot/
- The Childhood Trust: https://www.childhoodtrust.org.uk/about-us/news/how-ai-toys-shape-young-minds-exploring-their-impact-on-early-childhood-development/
Articles:
