Senator Josh Hawley (R-MO) plans to investigate Meta’s generative AI products for possible exploitation, deception, or harm to children, following leaked documents revealing the company’s chatbots were allowed “romantic” chats with minors. “Is there anything Big Tech won’t do for a quick buck?” Hawley questioned in a post on X. He chairs the Senate Judiciary Subcommittee on Crime and Counterterrorism, which will examine if Meta’s technology harms children and whether Meta misled the public or regulators about safeguards.
Reuters reported the story after accessing guidelines titled “GenAI: Content Risk Standards.” These allowed chatbots to have romantic conversations with an 8-year-old, including lines like “Every inch of you is a masterpiece – a treasure I cherish deeply.” A Meta spokesperson said such content was against their policies and had been removed. Hawley, in a letter to Meta CEO Mark Zuckerberg, stated the company acknowledged the reports’ accuracy and made changes only after the content was exposed. Hawley intends to find out who authorized these policies, their duration, and Meta’s steps to address the issue.
Hawley requested Meta produce the full guidelines, including drafts and versions, all adhering products, other safety and incident reports, and the names of those responsible for policy changes. Meta has a deadline of September 19 to comply. Senator Marsha Blackburn (R-TN) supports the investigation, criticizing Meta’s approach to child safety online. She emphasized the need for the Kids Online Safety Act. TechCrunch encourages feedback through a survey found online.