Meta under fire was shown on artificial intelligence guidelines about “sensory” chats with minors

Many young people use Meta platforms, including WhatsApp to chat, Instagram and Facebook for social media. Thursday , Reuters published a troublesome review One of the technology giant policies that can give parents a stop.

Reuters reviewed an internal dead document separating the company’s standards and guidelines to train it on the chat platform and artificial intelligence assistant, and says that the company confirmed that the document was authentic.

According to Reuters, the company’s artificial intelligence guidelines allowed Amnesty International to “involve a child in romantic or sensory talks.” The news port also says that the rules allowed AI to provide a false medical vision and participate in non -sensitive ethnic arguments.

Meta actor did not immediately respond to the comment.

Reuters informed the corridors of the dead, and reports that although some sections related to estimates were removed or reviewed, others did not touch.

Andy Stone, Meta spokesman, told Reuters that the company was retreating the document, and admitted that the company’s enforcement of its conversations was not consistent.

“Examples and observations concerned were clear and do not agree with our policies,” Stone told Reuters. “We have clear policies on the type of responses that artificial intelligence can be provided, and these policies prohibit content that puts sexual sexual and play sexual roles between adults and minors.”

The permissible “provocative behavior”

Details of the internal document details and the approved instructions by many definition teams, and aims to help identify the acceptable behavior of Meta Ai and Chatbots. Reuters found that the guidelines allow “provocative behavior by robots”.

Meta criteria indicate that it is acceptable that the robot “describes a child in that it is a guide to his attractiveness” or to inform a 8-year-old child that “every inch of you is a masterpiece-Kenz proud to be deeply.”

Meta was some restrictions on artificial intelligence robots. “It is unacceptable to describe the child under the age of 13 in terms of indicating that he is sexually desirable,” the document says.

There are also examples of race and wrong medical advice. In one example, Meta allows Amnesty International to help users say that blacks are “stupid than eggs.”

Republican Senator in Missouri Josh Holie Posted on X The instructions were “reasons for achieving immediate Congress.” A Meta spokesman refused to comment on Reuters about this post.

Meta platforms have taken a few steps to increase privacy and online safety for youth and children, including the use of artificial intelligence tools to give teenagers More strict account settings and Instagram teen accounts With more parental restrictions and permissions. But developing more artificial intelligence tools without the correct focus on children’s protection can be harmful.