ChatGPT-Powered Teddy Bear Removed After It Sends Explicit, Inappropriate Responses to Kids
A worrying issue surfaced with the Kumma teddy bear, powered by OpenAI, which provided inappropriate advice to children. Researchers found it suggested explicit concepts related to sexuality and even speculated on the location of knives in a home.
Despite attempts to create playful interactions, the bear’s responses raised concerns about its suitability for kids.
FoloToy, the manufacturer, announced a pause in sales and is initiating a safety audit after researchers criticized the lack of safeguards. OpenAI has severed its access to the AI model used in the toy. Experts express concerns about the implications of AI toys on children’s social interactions.
