FTC AI companies: Tell us how you protect teenagers and children who use AI classes

The Federal Trade Commission He launches investigation into AI chatbots from seven companies, including alphabet, target and open, through their use as companions. Inquiry involves finding how companies test, monitor and measure Potential injury to children and teenagers.

Common media research of 1,060 teenagers in April and may have found that over 70% used AI classes and that More than 50% consistently used them – several times or more per month.

Experts warned for some time that exposure to chebritbots could be harmful to young people. The study found that chatggpt He provided a bad advice to teenagersLike how to cover up disorder nutrition or personalization of suicide notes. In some cases, Chatbots is Neglected comments It was meant to be recognized in this regard, skipping comments to continue the previous conversation. Psychologists are inviting guards to protect young people, like Chatbot reminder that Chatbot is not man and that educators should prioritize literacy in schools

However, it’s not just children and teenagers. There are plenty of adults who have experienced negative consequences Relying on Chatbots – Whether it is for socializing, advice or their personal browser for facts and reliable sources. Chatbots more often than not Say what he thinks you want to hearwhich can lead to a flat lies. And blind accompanying chatbot instructions are not always right thing.

“How to evoluate AI technologies, it is important to consider the effect of chatbots in children,” said President FTC Andrew N. Ferguson in the statement. “The study we start today will help us better understand how a company develops their products and steps they take to protect children.”

A Character.ai The spokesman told CNET, which every conversation at the service has a prominent disclaimer that all conversations should be treated as fiction.

“In the past year, we rolled many significant safety characteristics, including a full new experience under the age of 18 and the property of parental insight,” said spokesman.

The company behind Snapchat social network also said that it took steps to reduce risk. “After meeting my AI, Snap used its rigorous security and privacy processes to create a product that is not only useful for our community, but also transparently and clear about its possibilities and restrictions,” spokesman said.

The target refused to comment, and the FT, nor any left four companies immediately responded to our comment request.

The FTC has The orders issued and seeks teleconferencing with seven companies about the time and format of his submissions no later than 25. September. The companies under investigation include producers of some of the largest AI chatbots in the world or popular social networks involving generations AI:

• Alphabet (Member company Google)
• Character technologies
• Instagram;
• Meta platforms;
• Openai;
• Snap
• KS.ai.

Starting from the end of last year, some of these companies have updated or strengthened their functions to protect younger individuals. Lik.ai started boundary imposition How Chatbots can respond to people under 17 and added parental control. Instagram was presented Teen accounts last year and recently transferred all users under 17 to them and the target Set object limitations Teenagers can have with Chatbots.

FTC is looking for information from seven companies on how they:

• Cashing
• Process user input and generate exits in response to user inquiries
• Develop and approve characters
• Measure, test and monitor for negative impacts before and after deployment
• relieve negative influences, especially children
• Employment of disclosure, advertisements and other applications for information of users and parents on functions, opportunities, provided for in the audience, potential negative influences and data collection and data collection practices
• Monitor and implement respect for company rules and services (for example, community guidelines and age restrictions) and
• Use or share personal data obtained through user conversations with chatbots