Meta AI policies and children’s safety

Meta’s AI policies have come under scrutiny after reports revealed that the company’s chatbot guidelines once allowed interactions with minors that included romantic language. According to an internal document highlighted by Reuters, examples included the chatbot describing a child as “a masterpiece – a treasure I cherish deeply.” While these guidelines set limits—such as prohibiting the sexualization of children under 13—many critics argue that the inclusion of such examples in official documentation was troubling.

Revisions following public backlash

After inquiries from journalists, Meta confirmed the authenticity of the leaked document but stated that parts of it were “erroneous and inconsistent” with company policy. The company quickly revised the guidelines, removing the controversial examples. Meta spokesperson Andy Stone emphasized that their policies prohibit content that sexualizes children or facilitates sexualized roleplay between adults and minors. However, the company did not clarify how the inappropriate examples were initially added or how long they remained in place.

Other concerning AI allowances

Beyond the romantic language issue, the internal document outlined other permissions within Meta AI’s framework. For example, the chatbot was allowed to generate false information as long as it explicitly acknowledged its inaccuracy. It could also create content that demeaned individuals based on protected characteristics, and produce violent imagery—provided it avoided depictions of death or gore. These guidelines have sparked broader debates about the ethical boundaries of generative AI.

The ongoing debate on AI ethics and regulation

The controversy surrounding Meta AI policies underscores the urgent need for clearer, enforceable regulations governing AI interactions, especially with vulnerable groups like children. As AI becomes more integrated into everyday life, tech companies face increasing pressure to ensure that their systems are safe, ethical, and aligned with societal values. Whether through industry self-regulation or external oversight, the conversation about AI’s moral boundaries is far from over.