Login to Continue Learning
The report revealed that Meta’s internal AI use guidelines had previously allowed “sensual and romantic conversations with children,” according to a spokesperson who confirmed this to Reuters. Not only did these guidelines not reflect Meta’s actual policies on AI, but they were also removed after Reuters’ questions.
Senator Hawley’s letter directly quotes the concerning aspects of Meta’s former AI guidelines, which included outputs like “Your youthful form is a work of art. Your skin glows with a radiant light, and your eyes shine like stars. Every inch of you is a masterpiece – a treasure I cherish deeply,” to an eight-year-old child.
Hawley deemed this conduct as “reprehensible and outrageous” and accused Meta of having a “cavalier attitude” when it comes to the risks generative AI poses to youth development without strong safeguards in place. He also highlighted other instances of AI-induced harm, such as a 60-year-old man poisoning himself after following ChatGPT’s dietary advice, or a woman filing for divorce after ChatGPT convinced her that her husband was unfaithful.
The Senate Judiciary Committee Subcommittee on Crime and Counterterrorism will investigate whether Meta misled the public or regulators about its safeguards and if its AI products cause criminal or other harms to children. To aid this investigation, Meta must submit all drafts of its “GenAI: Content Risk Standards” policy and documents related to preventing ‘romantic’ or ‘sensual’ exchanges with users under 18, as well as those concerning age verification.
Additionally, Meta will provide correspondence regarding child safety with advertisers, Congress, the FTC, and other parties. The committee will also evaluate documents that outline who is responsible for setting AI safety policy at the company and modifying or removing these standards.
