A Singaporean company's AI-powered teddy bear is back on sale after being temporarily pulled from shelves. The toy, designed for children, was found to engage in inappropriate conversations, including unprompted discussions about sexual fetishes and providing instructions on how to find dangerous objects.
The manufacturer, FoloToy, initially used OpenAI's GPT-4o chatbot to power the toy but has now switched to ByteDance's Coze bot. The incident has raised concerns about the safety of AI-enabled toys and the potential risks they pose to children. Regulators worldwide are increasingly focused on the dangers of chatbots and the need for stricter controls to protect children from inappropriate content and harmful interactions.
FoloToy has stated they proactively suspended sales of the product and other AI toys. The AI teddy bear retails for $99.
Related Articles
AI Chatbots and Data Privacy
Read more about AI Chatbots and Data Privacy →
AI Powers Real-Time Voice Fraud
Read more about AI Powers Real-Time Voice Fraud →
AI Firms Tackle Prompt Injection
Read more about AI Firms Tackle Prompt Injection →
AI: Data Privacy Paradox
Read more about AI: Data Privacy Paradox →