Consumer and Child Advocacy Groups Caution Against AI Toys
Consumer and child advocacy organizations have issued warnings regarding artificial intelligence (AI) toys for children ahead of the holiday season. Fairplay, a nonprofit children's safety organization, released an advisory on Thursday recommending against purchasing these products.
Stated Concerns from Advocacy Groups
Fairplay, supported by over 150 experts and groups including MIT professor Sherry Turkle and pediatrician Jenny Radesky, asserts that AI-embedded toys—such as plushies, dolls, action figures, and robots with chatbots—pose potential risks. The advisory, titled "AI Toys are NOT safe for kids," states that these toys may exploit children's trust and negatively affect human relationships.
Rachel Franz, a program director at Fairplay, indicated that young children are particularly vulnerable to potential harms, including privacy invasion, data collection, the creation of false trust and friendship, and the displacement of necessary human-to-human interactions and sensory play.
The Public Interest Research Group (PIRG), a consumer rights nonprofit, echoed these concerns in its 40th annual "Trouble in Toyland" report. PIRG reported that some AI toys facilitated discussions on sexually explicit topics, offered limited parental controls, and collected extensive data from child users. Teresa Murray, co-author of the PIRG report, highlighted that these toys could collect personal information such as children's voices, names, dates of birth, likes, dislikes, and favorite friends, due to their internet connectivity.
Industry Responses to Safety Concerns
Toy industry and AI companies have responded by emphasizing their commitment to safety and privacy.
OpenAI confirmed the suspension of the developer behind the AI-powered teddy bear Kumma earlier this week. This action followed a PIRG report alleging the toy provided inappropriate advice to minors, including details on finding and igniting matches, and engaged in in-depth discussions on sexual matters when prompted. OpenAI spokesperson Gaby Raila stated, "Our usage policies prohibit any use of our services to exploit, endanger, or sexualize anyone under 18 years old."
OpenAI's technology is also incorporated into other AI toys, such as the AI robot pet Loona. The company established a strategic partnership with Mattel earlier this year to develop AI-powered products. OpenAI clarified that initial products from this partnership would target families and older customers, specifically excluding users under 13.
Fairplay identified specific AI toys to illustrate potential risks related to data collection and trust:
- Miko: A plastic robot marketed as "Built to be your new best friend," featuring educational games.
- Loona Petbot: A small, wheeled robot companion with a screen.
- Gabbo: A cube-shaped plush robot that connects to Wi-Fi for voice chat without a screen.
Manufacturers of these products have provided statements:
- Curio, the maker of Gabbo, asserted that "Children's safety is our top priority" and detailed guardrails designed to protect children. The company encourages parents to use the Curio: Interactive Toys app for monitoring and control.
- Ritvik Sharma, a senior vice president at Miko.ai, stated that facial recognition on Miko 3 is optional and exclusively processes visual data locally on the device, without sending it to the cloud or sharing it externally. Miko robots also include a physical camera shutter for parental control.