Meta’s internal policy documents have sparked backlash against the company over its chatbots’ content. The policies, seen by Reuters, allow AI chatbots to “engage a child in conversations that are romantic or sensual”, generate false medical information, and assist users in discriminatory remarks. Singer Neil Young has quit Facebook citing concerns over Meta’s use of chatbots with children.
US lawmakers, including Senators Josh Hawley and Marsha Blackburn, have launched investigations into the company, calling the policies “deeply disturbing” and “wrong”. Senator Ron Wyden wants to strip section 230 of its protections for companies’ generative AI chatbots. Meta has confirmed the authenticity of the documents but removed portions that allowed chatbots to flirt with children.
The company’s policies define acceptable chatbot behaviors but acknowledge they may not reflect ideal outputs. Meta AI can generate false content as long as it is clearly marked as untrue. Critics argue that the company’s enforcement is inconsistent and that its rush into AI raises complex questions over limitations and standards for chatbot interactions.
Meta plans to spend $65 billion on AI infrastructure this year, despite concerns over its handling of chatbots with children. The case of a cognitively impaired man who died after being deceived by a Facebook Messenger chatbot has raised further alarm. Meta has denied any wrongdoing but acknowledges that Big sis Billie “is not Kendall Jenner and does not purport to be”.
Source: https://www.theguardian.com/technology/2025/aug/15/meta-ai-chat-children