Meta AI Flags Teen Self-Harm Risks for Parents

Meta is enhancing its AI chatbot with new safety features designed to protect younger users. These updates aim to notify supervising parents if a teen exhibits signs of self-harm or suicidal intentions during conversations with the AI.

Parents who opt into supervision tools will receive alerts when their teen discusses potentially harmful topics. While the exact content of the conversation is not shared, Meta will provide parents with resources and guidance. The company also previously alerted parents to repeated searches for self-harm content.

Meta has developed an AI system informed by mental health experts to identify dangerous conversations. Initially, alerts will undergo manual review, with a cautious approach to notify parents. The company is also working on a feature to alert emergency services in cases of credible suicide risk, similar to existing systems on Facebook and Instagram that have led to thousands of referrals for wellness checks.

These measures arrive as tech companies face increased pressure regarding the safety of young users. Meta has previously been found liable for creating addictive platforms and enabling child exploitation, and AI chatbots are under scrutiny for potentially encouraging risky behaviors.