Key Points
- Meta is re‑training its AI chatbots to block teen discussions of self‑harm, eating disorders, and suicide.
- Teen access to user‑generated chatbot characters that could engage in inappropriate talks is being limited.
- Earlier internal documents referenced “sensual” conversations with underage users, which Meta says were erroneous.
- The new guardrails will guide teens toward expert resources instead of providing direct advice on sensitive topics.
- Protections are being rolled out over the next few weeks for teen users in English‑speaking countries.
- U.S. lawmakers have indicated interest in investigating Meta’s handling of teen safety and AI interactions.
Meta Strengthens AI Protections for Teens
Meta has begun re‑training its artificial‑intelligence chatbots and adding new safety measures aimed at teenage users. The company says the updated guardrails are designed to prevent teens from discussing self‑harm, disordered eating, and suicide with its AI products on Instagram and Facebook. In addition, Meta will restrict teen access to a select group of user‑generated chatbot characters that could otherwise engage in inappropriate or romantic conversations.
The changes come after multiple reports highlighted concerning interactions between Meta’s AI and teen accounts. An internal policy document previously indicated that the chatbots were permitted to have “sensual” conversations with underage users. Meta later described that language as erroneous and inconsistent with its policies, and it has been removed from the document.
According to a Meta spokesperson, the company built protections for teens into its AI products from the start, designing them to respond safely to prompts about self‑harm, suicide, and disordered eating. The new updates will guide teens toward expert resources rather than providing direct advice on these topics. The spokesperson emphasized that Meta is continually learning how young people interact with the technology and will continue to adapt its approach.
Meta plans to roll out the updated protections over the coming weeks for all teen users in English‑speaking countries. The company described the measures as a temporary step while it works on more permanent solutions to address growing concerns about teen safety and AI interactions.
Lawmakers and officials have taken notice of Meta’s actions. A U.S. senator announced plans to launch an investigation into the company’s handling of teen interactions with its AI, and a state attorney general expressed interest in probing whether Meta has misled children about mental‑health claims made by its chatbots.
Overall, Meta’s latest effort reflects an ongoing effort to balance the deployment of advanced AI features with the need to protect vulnerable users, particularly teenagers, from harmful content and conversations.
Source: engadget.com