Meta is expanding safeguards for teen users with a new feature that alerts parents when their children's AI conversations suggest self-harm or suicide risk. When an underage user discusses these topics, Meta AI directs them to crisis helplines and encourages reaching out to an adult. Starting today, the company will also proactively notify supervising parents or guardians, sharing resources and advice.

Meta built a dedicated AI system to identify these conversations but manually reviews all flagged chats before sending alerts to avoid false alarms. The system errs on the side of caution, notifying parents even when intent seems ambiguous. This feature is live for Instagram parental supervision in the US, UK, Australia, and Canada, with global rollout by year-end.

The company is also developing the ability to contact emergency services if a user's conversation suggests imminent risk, regardless of age. In May, OpenAI launched a similar Trusted Contact feature for ChatGPT. Additionally, Meta's Limited Content setting now extends to AI interactions, filtering sensitive topics when enabled by parents.

For immediate help, contact the National Suicide Prevention Lifeline at 1-800-273-8255, available 24/7 with online chat option.