Emily Goodacre, a researcher at The University of Cambridge has raised concerns over the use of AI in children’s toys.
As AI develops, it only makes logical sense that businesses look to implement AI in all manners of everyday goods. This new development has seen a new selection of AI toys, such as teddy bears that run on Open AI’s Chat GPT-4o.
These teddy bears work by letting children pick a theme for their AI bear, but there have been instances of these bears blurting out unrestricted content: The U.S. Public Interest Research Group Education Fund (PIRG) found AI toys to express disturbing material, one teddy bear going into explicit detail over sexual kinks, while another glorified instances of dying in battle from Norse mythology. Upon being approached for comment, PIRG said ‘AI toys rely on tech that is very new, basically unregulated and that […] we’ve already found comes with real concerns. It’s clear there are a lot of questions around chatbots and kids we need to answer before the market takes off.’
This caused FoloToy to withdraw its ‘Kumma’ bear and other AI-enabled toys, with an internal safety audit occurring and OpenAI suspending the developer for violating its policy regarding under 18s.
Aside from the current issues with safety, experts argue that even when working perfectly, ethically these AI toys should be discussed before children are given free reign.
Emily Goodacre, a researcher with the Centre for Research on Play in Education, Development and Learning at the University of Cambridge, has been studying AI-powered toys and their effect on long-term development. During an interview with Yahoo, Goodacre noted how teachers were concerned about the lack of true human social interaction, yet consistently made clear that the evidence was not conclusive either way, and that there were societal questions we should be asking.
She also highlighted the privacy concerns as the transcripts are sent to the parent’s phone. Goodacre argues that as a society we should be discussing the increased technology (from Alexa to Siri to these new toys) we have in our rooms and the decrease in privacy, and especially be considering the child’s perspective.
These concerns with AI come against the backdrop of incidents such as AI therapy encouraging people to die by suicide and AI self-driving cars crashing into walls.
This clear danger with AI is leading legislators in the US to consider pushing for AI regulation bills, however the UK government seems yet to have considered rushing its planned AI bills.
However, these AI teddy bears do not fit the standards for chatbots, and concerns about the extent to which legislation can protect against this new development. For example, legislation may require AI to remind the user that it is not human, or if dangerous topics come up, signpost the user to appropriate resources. These suggestions might be awkward to implement in a toy and calls for creative law-making to ensure all situations are covered.
It remains to be seen whether AI has a place in the market of children’s toys, and these issues certainly need to be addressed.
Featured Image Credit: NBC Connecticut
