Meta said on July 17 it would alert parents using Instagram parental supervision in the US, the UK, Australia, and Canada if a teen’s chat with Meta AI suggested possible suicide or self-harm risk.

The company said it used signals developed with experts and manually reviewed flagged chats before sending alerts. Meta plans to expand the alerts to parents globally by year-end.

Meta said its AI assistant already directs teens to crisis helplines and trusted adults. It is also building a feature to contact emergency services when a chat suggests an adult or teen may be at imminent risk of taking their own life.

Meta also said Instagram’s stricter Limited Content setting will apply to Meta AI chats for teens, widening the range of prompts the chatbot will decline.

Meta said the setting is more restrictive than its default 13+ standard and is meant to further limit age-inappropriate material. It uses a movie-ratings-inspired framework designed for teens.

Meta also said teen AI protections apply to users who entered a teen birthday or whom its systems infer are teens.

🔗 Source: Meta

Recent Meta developments