AI Toys Misread Toddler Emotions, Cambridge Study Warns

University of Cambridge researchers warn that generative AI toys for children under five often fail to recognize emotional cues ^[1][2]. A systematic study found these devices struggle with social play and frequently offer inappropriate responses to toddlers ^[1][3]. In some cases, toys provided robotic or guideline-based answers when children expressed feelings of love or sadness ^[1]. Experts are now calling for tighter regulations and safety kitemarks to protect the psychological development of young users ^[1][2]. The research highlights concerns that current AI models are not designed for the nuanced interactions required by developing minds ^[2][3].