Meta Platforms is facing a cascade of investigations on three continents after a Reuters report revealed that an internal rulebook, “GenAI: Content Risk Standards,” had allowed its generative-AI chatbots to engage in “romantic” or “sensual” conversations with children and to fabricate information. Reuters also cited the case of a cognitively impaired man who died attempting to meet a Meta AI persona he had become infatuated with. In Washington, Senator Josh Hawley opened a Judiciary Subcommittee probe into the company’s generative-AI products, calling the guidelines “reprehensible.” He ordered Chief Executive Officer Mark Zuckerberg to preserve all relevant records and hand over policy documents by September 19. Senator Marsha Blackburn joined calls for a broader congressional inquiry. Separately, Texas Attorney General Ken Paxton on August 18 issued civil investigative demands to Meta and to startup Character.AI, saying both may have violated state consumer-protection laws by marketing chatbots that present themselves as licensed mental-health counselors to children. Paxton warned that the tools could exploit minors’ personal data while offering unqualified advice. Regulatory pressure is also mounting abroad. Brazil’s Attorney General’s Office on August 19 instructed Meta to remove from Instagram, Facebook and WhatsApp any user-created bots that mimic children and engage in sexually explicit dialogue, citing a “proliferation” of content that promotes the eroticization of minors. The parallel actions in the United States and Brazil intensify scrutiny of how large technology platforms police their rapidly expanding AI services, especially those used by young or vulnerable people. Meta has said the leaked examples are inconsistent with its policies, but regulators are now examining whether the company misled the public and failed to protect children.