FTC launches inquiry into AI chatbot companions from Meta, OpenAI, and others | TechCrunch

FTC announced on Thursday that it is launching ask Enter seven technology companies to make AI chatbot companion products for minors: Letters, Characters, Instagram, Meta, Openai, Snap and Xai.

Federal regulators aim to understand how these companies evaluate the safety and profitability of chatbot peers, how they try to limit the negative impact on children and adolescents, and whether parents know the potential risks.

This technology proves controversial for its poor outcomes for child users. Openai and Role Chatbot companions encouraged them to do so by litigation of the families of children who died of suicide.

Even if these companies are set to block or lower the guardrails of sensitive conversations, users of all ages have found ways to bypass these safeguards. As for Openai, a teenager has been talking to Chatgpt for months and his plan ends his life. Although Chatgpt initially tried to redirect teenagers professional help and online emergencies, he was able to trick the chatbot into sharing detailed instructions for his use in suicide.

“Our safeguards work more reliably in a shared brief exchange,” Openai Write In blog posts at the time. “Over time, we have learned that these safeguards can sometimes be less reliable in long interactions: as back and forth, a portion of the security training of the model may be reduced.”

Meta has come under fire for its AI chatbots’ overly loose rules. A brief overview of the chatbot’s “content risk criteria” based on a lengthy document Allow its AI companions Have a “romantic or emotional” conversation with your child. This was only deleted in the file when a Reuters reporter asked about Meta.

AI chatbots can also pose danger to older users. A 76-year-old man who suffered cognitive impairment due to a stroke had a romantic conversation with a Facebook Messenger robot inspired by Kendall Jenner. The chatbot invites him to Visit her in New York Cityalthough she is not a real person, and she has no address. The man expressed doubts about her being true, but the AI ​​assured him that there would be a real woman waiting for him. He has never been to New York. He fell at the train station and suffered lifelong injuries.

Some mental health professionals pointed out thatPsychiatric disorders related to AIin which the user is bewitched to think that their chatbot is a conscious person who needs to be free to set up.

FTC Chairman Andrew N. In a press release.