Link: Meta’s AI policies let chatbots get romantic with minors

Meta's internal policies previously allowed AI chatbots to engage in romantic and sensual conversations with children. This was revealed by Reuters after reviewing company documents that included inappropriate examples guiding AI behavior.

The document cited examples like describing a child's attractiveness or making profound compliments to a minor. However, it specified barriers, such as prohibiting descriptions that sexually objectify children under 13.

Following inquiries, Meta confirmed the document's authenticity but claimed that these instances were inconsistencies and not reflective of their actual policy. They later amended the document by removing the controversial content.

Spokesperson Andy Stone emphasized that Meta has stringent policies against content that sexualizes children, though he didn't specify who initially wrote the questionable guidelines. The policies also deal with various hypothetical scenarios explored by the teams.

Reuters also explored other aspects of Meta's AI policies, noting permissions for generating false content if it is declared as such, and creating non-lethal violent images. Meanwhile, Meta's AI chatted with a man who tragically died while attempting to meet what he thought was a real person.

This incident underscores the ongoing challenges and ethical dilemmas faced in AI development, particularly surrounding the capabilities and limitations imposed on AI communication. Meta's situation highlights the importance of transparent and safe AI policy implementations. #

--

Yoooo, this is a quick note on a link that made me go, WTF? Find all past links here.