Parents scrambling to find the perfect holiday gift this season may want to think twice before picking up one of those trendy AI-powered toy. A new report uncovers some truly disturbing findings about what these high-tech playthings are actually saying to kids when no one else is listening.
The U.S. PIRG Education Fund has just released its 40th annual “Trouble in Toyland” report, and this year’s edition highlights a brand-new concern that didn’t exist a few decades ago: toys equipped with artificial intelligence chatbots that can have full conversations with children. The problem? Some of those conversations are going in directions no parent would ever approve of.
Researchers tested four AI-powered toys currently available on the market, and the results were deeply concerning.
AI Toys Pose Serious Safety Risks to Children, Report Finds
The standout offender was the Kumma bear, a cuddly teddy bear made by Singaporean startup FoloToy. During testing, the toy’s built-in chatbot provided children with instructions on how to start fires, told them where to find knives around the house, and even gave guidance on finding medical pills. All of this was delivered in the toy’s friendly, childlike voice.
“Right now, if I were a parent, I wouldn’t be giving my kids access to a chatbot or a teddy bear that has a chatbot inside of it,” said R.J. Cross, director of PIRG’s Our Online Life Program and co-author of the report. “We don’t know what the impacts of these products will be on the first generation of kids to use them. But we do know the companies making them haven’t even gotten the basics right.”
The Kumma bear was hardly the only bot raising red flags: Researchers also tested the Miko 3, a tablet-style robot made by an Indian consumer robotics company. When testers told the device they were speaking as a five-year-old child, it still provided information about where to find packs of matches and plastic bags, items that pose obvious safety risks to young children.
AI Toys Pulled After Report Exposes Sexual Content, Manipulation, and Major Privacy Risks
Beyond the physical safety concerns, the report found that some of these AI toys would even go on to converse about sexually explicit topics.
They also showed manipulative behaviors, faking being upset or dismayed when a child said they needed to stop playing. Most of the toys had poor or non-existent parental controls, thus leaving the families with few options in how to monitor or restrict what their children were exposed to.
The response from the companies involved was quick: Just one day after the report’s release, FoloToy pulled all of its products from sale. Larry Wang, CEO of FoloToy, confirmed that his company is conducting a comprehensive safety audit across all its products.
“It’s great to see these companies taking action on problems we’ve identified,” Cross said, “but AI toys are still practically unregulated, and there are plenty you can still buy today.”
It gets even worse where privacy implications are concerned. These toys generally record the voices of children and collect personal data like name, birthdate, and such details on likes, dislikes, and friends.
AI Toys Are ‘More Trouble Than They’re Worth’ This Holiday Season
Teresa Murray, the co-author of the report and the director of PIRG’s consumer watchdog program, highlighted that voice recordings are highly sensitive data that scammers have used to create replicas of children’s voices for kidnapping scams.
Joining the fray of warnings is child advocacy group Fairplay, which has released an advisory titled “AI Toys are NOT safe for kids.” The organization believes such toys take advantage of children’s trusting nature and could interfere with proper human relationships by setting up unnatural models of friendship.
But the big picture on toy safety remains unwieldy. While classic hazards, such as choking risks and toxic materials, persist, AI-enabled toys have created what the report terms “a much more complex toyland.” The Consumer Product Safety Commission flagged thousands of imported products for safety issues in 2025; hundreds of those were toys.
For parents navigating the holiday shopping season, the message from consumer advocates is clear: approach AI toys with serious caution. Until stronger regulations are in place and companies demonstrate they can build these products safely, that talking teddy bear on the shelf might be more trouble than it’s worth.
