Among the top concerns surrounding the toy was a common refrain regarding safety, in discussions about generative AI tools interacting with minors. Age-related safeguards on these tools are not always effective and users can be exposed to mature, dangerous or harmful content. This was happening with the AI teddy bear, called Kumma, with reports that it was discussing topics like knives, sexual subjects and giving instructions on how to light matches.
Is there an AI-powered teddy bear?
Answer: Yes, but even its creator has decided it’s not a good idea.
Shutterstock/Ann in the uk
If you think a teddy bear powered by ChatGPT is a bad idea, you’re not alone. Toymaker FoloToy is pulling its ChatGPT-enabled teddy bear over a variety of concerns.
Among the top concerns surrounding the toy was a common refrain regarding safety, in discussions about generative AI tools interacting with minors. Age-related safeguards on these tools are not always effective and users can be exposed to mature, dangerous or harmful content. This was happening with the AI teddy bear, called Kumma, with reports that it was discussing topics like knives, sexual subjects and giving instructions on how to light matches.
“FoloToy has decided to temporarily suspend sales of the affected product and begin a comprehensive internal safety audit,” FoloToy Marketing Director Hugo Wu said. “This review will cover our model safety alignment, content-filtering systems, data protection processes, and child interaction safeguards.”
Among the top concerns surrounding the toy was a common refrain regarding safety, in discussions about generative AI tools interacting with minors. Age-related safeguards on these tools are not always effective and users can be exposed to mature, dangerous or harmful content. This was happening with the AI teddy bear, called Kumma, with reports that it was discussing topics like knives, sexual subjects and giving instructions on how to light matches.
