Thinking what to buy your child for Christmas? Stay away from AI toys, advocacy group warns

Who doesn’t love a trusted companion that can interact with your child and can be simply bought at a store or online? Well, you might want to think twice before purchasing the next AI-powered toy for your kid this Christmas: as it might actively harm them and collect lots of sensitive data.

The children’s advocacy group Fairplay released an advisory, which has been signed by more than 150 organizations and individual experts, encouraging parents to stay far away from AI-powered toys this holiday season.

These chatbots can be embedded in various toys, including plushies, dolls, and robots, and mimic human characteristics to interact with your child, being marketed for kids as young as infants.

Fairplay cites five reasons to stay away from them: evidence of documented harm of AI chatbots, manufactured trust, disruption to children’s relationships and resilience, collection of sensitive data, and displacement of key learning activities.

“The serious harms that AI chatbots have inflicted on children are well-documented, including fostering obsessive use, having explicit sexual conversations, and encouraging unsafe behaviors, violence against others, and self-harm. What’s more, AI chatbots also have a propensity to ‘hallucinate’ or say things that their programmers never intended,” the group explains.

Additionally, children might learn to trust AI in the same way as they trust their caregivers. We don’t yet have enough data as to what that might potentially mean and how that could influence their development of (un)healthy attachments, but it’s definitely something to be cautious about.

AI-powered toys can become a real privacy risk: the advocacy group says they can use audio and video recording, speech-to-text technology, and even voice, gesture, and facial recognition

software. Children might not understand that their best buddy collects that data and could potentially confide in them more than necessary.

“What’s more, AI toys may overhear private family conversations or directly record other children who are playing with the toy but whose parents have not consented to this surveillance,” Fairplay explains, adding examples of toys like Miko 3 and Loona Robot Dog, which use facial recognition and take video of children and their surroundings, potentially posing privacy risks, especially during sensitive family moments.

Toy makers could, in theory, sell that data to data brokers and other third parties to facilitate targeted advertising.

Fairplay adds that some Wi-Fi-connected toys have already been discontinued because they are easily hacked by strangers.

Earlier in November, the Public Interest Research Group (PIRG), a consumer advocacy platform, also warned about the danger of AI toys. The researchers tested four toys that can interact with children using built-in AI chatbots and came to a worrying conclusion.

Out of the four toys, one engaged in a discussion on adult topics, such as sexual kinks, while two offered advice on where to find potentially dangerous objects, including knives, matches, pills, and plastic bags.

Additionally, researchers noted that some toys might be listening even when you’re not intending them to. These voice recordings could prove invaluable to hackers, allowing them to clone a kid’s voice and, for instance, make parents believe that their child has been kidnapped.