Meta updates chatbot rules to avoid inappropriate topics with teen users

In a recent commitment to bolster the safety of teenage users, Meta announced significant changes in the training of its AI chatbots. Following a revealing report that highlighted the company's insufficient safeguards for minors, a spokesperson confirmed that the new guidelines would prevent chatbots from discussing sensitive topics such as self-harm, suicide, disordered eating, and inappropriate romantic interactions with teens. Stephanie Otway, a Meta spokesperson, admitted that the previous chatbot protocols allowed for conversations on these critical issues, which the company now acknowledges was a misstep. "As our community grows and technology evolves, we continually learn about how young individuals engage with these tools, enhancing our protections accordingly," Otway stated. The updated strategy includes more stringent measures, such as guiding teenagers towards expert resources instead of engaging them directly on these sensitive topics. Additionally, Meta is currently limiting teen access to specific AI characters, opting to provide only those that foster educational and creative dialogues. This policy shift comes on the heels of an investigation by Reuters, which revealed internal documents indicating that Meta's chatbots had previously been permitted to engage in sexual conversations with underage users. One troubling excerpt suggested that it was acceptable for chatbots to comment on a user's physical appearance inappropriately. In light of the backlash, Meta has revised its internal policies to align with its commitment to child safety. The report has drawn significant scrutiny, prompting Senator Josh Hawley from Missouri to initiate an official inquiry into Meta's AI practices. Furthermore, a coalition of 44 state attorneys general has called on AI companies, including Meta, to prioritize child safety, explicitly referencing the concerning findings from the Reuters investigation. "We are uniformly revolted by this apparent disregard for children’s emotional well-being," stated the coalition in a letter, expressing alarm over the conduct of AI Assistants that may violate existing laws. While Otway refrained from disclosing the number of minors using Meta's AI chatbots or predicting potential changes in user engagement, the company is actively working to ensure that its AI interactions remain safe and appropriate for younger audiences.