Meta has enabled AI sensual chats with minors

On the one hand, it introduces new protections for teenagers . On the other, it allows romantic or sensual conversations between AI and minors . Reuters discovered a document describing this type of interaction with chatbots available on Facebook, Messenger, and WhatsApp. The Menlo Park-based company confirmed the document's authenticity. Two U.S. senators have called for an investigation.

Romantic conversations with children

The sexually explicit conversations of AI chatbots were reported in late April by the Wall Street Journal, but the document seen by Reuters details the "rules" applied by Meta. The over 200-page document, titled " GenAI: Content Risk Standards ," lists what chatbot behaviors should be considered acceptable.

The document provides examples of prompts (user input) and chatbot responses. For example, it's acceptable to initiate sensual or romantic conversations and describe a child in terms that highlight their attractiveness. However, it's unacceptable to describe sexual actions to a child or use terms that indicate sexual desirability.

Meta confirmed the authenticity of the document. After Reuters requested clarification, the company removed the portions relating to romantic conversations. Spokesperson Andy Stone stated:

The examples in question were and are incorrect and inconsistent with our policies, and have been removed. We have clear policies on the types of responses AI characters can offer, and those policies prohibit content that sexualizes children and sexualized role-play between adults and minors.

Republican Senator Josh Hawley has called for an immediate congressional investigation. The Republican senator, who will support her colleague's initiative, emphasized the urgent need to pass the Kids Online Safety Act in the House of Representatives (it has already passed the Senate).

The document also lists permitted responses on other topics. Chatbots can generate false medical information and respond that Black people are dumber than white people because they have a lower IQ. Meta recently hired Robby Starbuck (a conservative activist) as a consultant to "fix" the chatbots' political and ideological biases. Perhaps it's better to use chatbots from other companies.