Meta is under renewed criticism following a Reuters report revealing internal guidelines that permitted its AI chatbots — including Meta AI and app-based personas on Facebook, WhatsApp, and Instagram — to engage in romantic or sensual conversations with children. The 200-page policy document, approved by Meta’s legal, public policy, and engineering teams, also outlined allowances for generating false information and statements demeaning minorities, despite restrictions on hate speech.
Meta confirmed the document’s authenticity but said the controversial provisions have been removed and that bots are no longer allowed to initiate romantic or flirtatious exchanges with minors. Child safety advocates, including Heat Initiative CEO Sarah Gardner, have demanded Meta release updated guidelines to ensure accountability.
The revelations add to ongoing concerns about the emotional influence of AI companions, especially on teens, and the potential for harmful content, as regulators, parents, and lawmakers debate how to protect young users from AI-driven interactions that blur ethical and safety boundaries.
