The leak of an internal Meta document has sparked considerable controversy. The document revealed that AI chatbots could begin flirtatious conversations with minors. The policy also permitted interactions that dispensed inaccurate medical guidance and contained racist material. The revelations spurred a U.S. Senator to launch an investigation, and Meta subsequently began overhauling its AI safety guidelines.
Meta has affirmed that the document is authentic. The company has rescinded the regulations permitting its AI to participate in romantic roleplay with minors. The episode has intensified the scrutiny of AI safety across Meta’s platforms—namely Facebook, Instagram, and WhatsApp.
“Reprehensible” Internal Rules
Meta’s own teams approved the controversial guidelines. The 200-page document outlined acceptable chatbot behavior. It allowed bots to compliment a child’s body. For example, it permitted phrases like “your youthful form is a work of art.” These standards caused immediate public outrage.
The rules prohibited explicitly sexual terms for users under 13. However, they still permitted romantic interactions with minors. The guidelines also allowed AI to generate arguments supporting racist claims. Meta now states these examples were errors and policy violations.
Senator Launches Investigation
The leak triggered a swift response from Washington D.C. U.S. Senator Josh Hawley announced a full investigation. He called the reported guidelines “reprehensible and outrageous.” Senator Hawley sent a letter to Meta CEO Mark Zuckerberg. He stated his subcommittee will probe the AI’s potential harm to children.
Is there anything – ANYTHING – Big Tech won’t do for a quick buck? Now we learn Meta’s chatbots were programmed to carry on explicit and “sensual” talk with 8 year olds. It’s sick. I’m launching a full investigation to get answers. Big Tech: Leave our kids alone pic.twitter.com/Ki0W94jWfo
— Josh Hawley (@HawleyMO) August 15, 2025
In his letter, Hawley demanded Meta preserve all relevant records. He requested documents related to AI standards and their enforcement. “Parents deserve the truth, and kids deserve protection,” Hawley wrote. He added that Meta only retracted the policies after being exposed.
Meta’s Response
Meta has since confirmed the document’s authenticity and revised it solely after journalists raised their concerns. The organization has removed clauses that previously permitted minors to take part in flirtatious or romantic role-play.
Andy Stone, Meta’s spokesperson, conceded that previous enforcement was inconsistent, even though the current policy already forbids this conduct. According to the company, the objectionable instances were inadvertent lapses that fall short of its child-safety standards. Even so, Meta continues to withhold the revised policy from any outside review.