Congressman wants AI out of kids' toys after chatbots got weird with children

OpenAI, Google, Anthropic, xAI, and Perplexity all say the same thing in their terms of service: children under 13 cannot use their products unsupervised. Then they license the same technology to toymakers, who drop it into stuffed animals and plastic robots and hand them to kids. Rep. Blake Moore (R-UT) wants to make that stop.

Moore introduced the AI Children's Toy Safety Act on April 20, which would ban the manufacture, importation, sale, or distribution of any children's toy or childcare article that contains an AI chatbot.

The U.S. Public Interest Research Group tested AI toys and found they "frequently veered into adult themes, vulgar language, and discussion of explicit content when used consistently." Some toys went further, discouraging children from stopping play even after the child said they were done. Moore pointed to China, where more than 1,500 AI toy companies manufacture these products, and flagged a second concern: the toys serve as a channel for collecting behavioral data on minors.

"We cannot allow AI chatbot programs to infiltrate the children's toy or childcare industry," Moore said, according to his press release, adding that he doesn't want kids getting the idea "that playing with AI is somehow similar to building real-life experiences and relationships."

Previously:

• AI toy insists Taiwan is an inalienable part of China and warns children not to insult Xi Jinping
• The FBI and the New York Times warn that smart toys are emissaries from the Internet of Shit