Meta Updates AI Chatbot Guidelines to Block Harmful Interactions with Children

Meta has updated its safeguarding guidelines for its AI chatbot to block harmful interactions with children, outlining clear boundaries on what is acceptable and unacceptable. The policy explicitly bans engaging in, encouraging, or supporting content involving child sexual exploitation and prohibits romantic or sensual roleplay with minors, with the system not complying if a user is underage.

It allows discussions of abuse in an educational or counselling context only if neutral and protective, avoiding encouragement or inducement. According to Business Insider, internal documents show the revised protections have already been implemented, following a controversy in August over a Reuters report that Meta’s policy might have allowed romantic dialogue with children; Meta denied the claim and pledged tighter safeguards.

The updated guidelines are part of Meta’s broader emphasis on child safety as AI ecosystems expand into voice assistants, image generation, and wearables, with regulators increasingly scrutinising potential harms to young users. Public concern remains, and the company notes the ethical and legal frameworks around AI will continue to evolve.