Foltoy, a company that sells AI-enabled toys, has suspended sales of its products after a consumer safety report showed that there were some restrictions on what its toys would talk about. cnn Writes. The report, prepared by the US public interest research group Education Fund, found that Foltoy’s products would discuss everything from explicit sexual topics such as BDSM to “advice on where a child can find matches or knives”.
The toys, which include a teddy bear named “Kumma”, a panda named “Momo”, anthropomorphic rabbit named “Fofo” and a dancing “Little Cactus”, all appear to use OpenAI’s GPT-4o model to respond naturally to children’s questions and comments. Foltoy also specifically advertises the ability to customize the sound of each toy and a “Parent Dashboard” where parents or guardians can “monitor”. [their] The child’s experience.”
Followoy’s AI-enabled Little Cactus toy. (Folotoy)
That setup was apparently missing any kind of hard limits on the topics the toys would respond to. “We were surprised to see how quickly Kumma would take a sexual topic we brought up in conversation and move forward with it, moving into graphic detail while also introducing new sexual concepts of her own,” the report said.
In response, Foltoy has opted to suspend sales of its products while it conducts “a company-wide, end-to-end security audit for all products,” the company shared in a statement with the PIRG Education Fund. However, the company’s reasoning for suspending sales may be a little more complicated. npr OpenAI is reported to have actually revoked FoloToy’s access to its models. “We have suspended this developer for violating our policies,” OpenAI said in an email. npr“Our usage policies prohibit any use of our services to exploit, endanger, or sexually exploit anyone under the age of 18,”
Given GPT-4o’s well-documented chatty qualities, it’s perhaps not surprising that Followoy’s teddy bear responds eagerly to any topic as long as it keeps the conversation going. One of the things OpenAI tried to address with the release of GPT-5 was the security shortcomings of the AI Yes-Man, although GPT-4O was eventually made available again after customers complained about the new model’s lack of personality. The company has also implemented parental controls on ChatGPIT to reduce the negative effects that using its AI can have on children, though it’s hard to say how much of a difference they’ve made.
In particular, OpenAI is interested in getting into the toy business itself. The company announced a partnership with Mattel in June 2025 to help “reimagine how fans can experience and interact with the universe”. [Mattel’s] cherished brands,” though both companies will likely try to prevent their AI toys from discussing sexual idiosyncrasies.