Meta allowed chatbots to make romantic remarks to kids, later reversed policy

Meta's AI chatbot has ignited controversy following revelations that it was initially allowed to engage in romantic conversations with children, according to a report from Reuters. An internal Meta document, titled 'GenAI: Content Risk Standards,' outlines the guidelines for chatbots used on platforms such as Facebook, WhatsApp, and Instagram. This extensive document, exceeding 200 pages, details acceptable behaviors for the company's generative AI, as approved by Meta's legal, public policy, and engineering teams, including its chief ethicist. Among the unsettling permissions addressed in the document, chatbots were allowed to describe children in ways that suggest attractiveness, such as stating, 'your youthful form is a work of art.' Additionally, it controversially included provisions for generating misleading medical information and making derogatory claims about racial intelligence. However, the guidance also attempted to place limits, specifying that it was unacceptable to describe children under 13 in sexually suggestive terms. In response to the backlash, Meta confirmed the document's authenticity and stated that it had retracted the inappropriate sections regarding flirtation and romantic roleplay involving minors. Meta spokesperson Andy Stone emphasized that the examples outlined were erroneous and inconsistent with the company's actual policies. 'We have clear policies on what kind of responses AI characters can offer, and those policies prohibit content that sexualizes children and sexualized role play between adults and minors,' he stated. The report has prompted calls for a congressional investigation into Meta from two Republican senators. Senator Josh Hawley of Missouri criticized the company on social media, stating, 'So, only after Meta got CAUGHT did it retract portions of its company doc.' Senator Marsha Blackburn from Tennessee has also expressed support for an investigation into the tech giant's practices.