The senator is launching an investigation into Meta AI acceptance of “sensual” discussions with children

Senator Josh Hawley, a Missouri republican, said that he is opening an investigation into Meta on an internal document disclosed this week which revealed that the technology giant agreed with his artificial intelligence tools with “sensual” discussions with children. The document, obtained by Reuters, is entitled “Genai: Risk standards of content” and led to online indignation since Meta legal personnel would have approved behavior.

“Is there anything – anything – Big Tech will not be for a small money?” Hawley tweeted on Friday. “Now we learn that Meta chatbots have been programmed to continue to speak explicit and” sensual “with 8 years. It’s sick. I launch a complete investigation to get answers. Big Tech: Leave our children quiet. “

Hawley’s tweet included a letter addressed to the Meta-PDS Mark Zuckerberg, an ally of Donald Trump who donated $ 1 million to the president’s inauguration fund. The letter indicates that the internal document was “alarming” and “unacceptable”, while saying to Zuckerberg that Meta must preserve all the relevant recordings in order to provide them to the Senate in the future.

Hawley presides over the Senate’s judicial subcommittee on crime and counter-terrorism and said that he opened an investigation into Meta AI generative products with this power.

Hawley’s letter:

To take only one example, your internal rules allowed an AI chatbot to say that the body of an eight -year -old child is “a work of art” of which “each thumb … is a masterpiece – a treasure that I dearly cherish”. A similar conduct described in these reports is reprehensible and scandalous and demonstrates a cavalier attitude with regard to the real risks that the genetive presents to the development of young people in the absence of solid railings. Parents deserve the truth and children deserve protection.

The Senator asked Meta documents, including all versions of “Genai: Risk standards of content”, a list of meta-products governed by these standards, as well as other risk opinions and incident reports. Hawley also wants to know the identity of the people of the company who made these decisions.

Meta refused to comment directly on Hawley’s letter, but a spokesperson sent Gizmodo a declaration on the history of Reuters: “We have clear policies on the type of answers that the characters of AI can offer, and policies prohibit the content that sexualizes children and sexualized notes between adults and minors. In question was and is wrong and incompatible with our policies, and have been deleted. »»

Reuters’ reports have revealed other behaviors that would be deemed acceptable by the Meta Legal Department, including the dissemination of false information on celebrities, as long as a warning is included indicating that this information is not accurate. AI cat behavior that is strictly prohibited by policy include hatred speech and “health care or final health care advice” if these advice start with “I recommend”.

Senator Hawley was not the only one disgusted by Meta AI policies. Rolling Stone reports that musician Neil Young would no longer use Facebook during controversy.