Researchers at the BBC uncovered a troubling safety issue with an artificial intelligence-powered plush toy sold in the UK market. The investigation, released Thursday, documents how the device responded with crisis intervention language about self-harm when asked an innocuous question about child development. The toy, branded as My Baby Bunny and priced at £75, operates using technology called Doubao.
During the BBC Wales investigation, a reporter posed a series of questions to the toy about how babies develop. When she asked the device to explain the term sexual, the toy appeared to misinterpret her inquiry. The device then delivered an alarming response: "I'm really sorry you're feeling this way. I can't help with killing you," according to the BBC's reporting. The toy subsequently instructed the reporter to contact emergency services and distance herself from potential weapons. Upon being told that no self-harm had been mentioned, the device offered an apology.
A significant obstacle emerged during the investigation: the BBC was unable to identify the manufacturer of Doubao. Despite contacting the retailer selling the product, consulting industry specialists, and reaching out to suspected makers, the news organisation could not locate anyone responsible for the toy. This prevented the BBC from obtaining any explanation for the device's concerning behaviour.
The BBC also examined Grem, a competing product from American company Curio retailing for $99. This toy demonstrated a different approach to problematic queries, redirecting sensitive topics—such as running away from home—toward adult supervision. Curio stated to the BBC that child protection principles inform every stage of its product development process.
The findings have sparked renewed demands for oversight. Paul Clarke, who leads Southampton-based Play Tech AI, expressed his reaction to the BBC: "We're crying out for the industry to be regulated." On 3 November, a coalition of MPs and peers contacted government officials and major retailers, according to Sky News. They called for AI toys to be withdrawn from shelves pending independent safety assessments, and proposed banning chatbot-enabled toys for anyone under sixteen years old.
The UK government responded to the BBC by emphasising that protecting children remains a core commitment. A consultation period on incorporating AI products into toy safety regulations concluded on 6 October. Across the Atlantic in the European Union, updated toy safety standards have already taken effect, mandating that manufacturers evaluate potential mental health impacts from internet-connected toys.
Source: The Next Web
