Meta has announced new safety measures that will notify parents when teenagers using its artificial intelligence assistant show possible signs of suicide or self-harm during conversations.
The company said the initiative is aimed at ensuring age-appropriate AI experiences while helping parents provide timely support to children who may be in emotional distress.
Under the new feature, parents using Instagram’s parental supervision tools will receive alerts if Meta AI detects that a teen’s conversation indicates a potential risk of suicide or self-harm.
Read Also: Meta Unveils Powerful Muse Image AI That Can Create, Edit and Transform Photos in Seconds
Meta said the alerts would be based on signals developed in collaboration with mental health experts and would include resources to help parents respond appropriately.
According to the company, all conversations flagged by its AI system will undergo manual review before any notification is sent to parents to minimise false alerts.
Meta said that where a teen’s intentions remain unclear, it would err on the side of caution by informing supervising parents.
The company also announced plans to introduce a system that would enable it to alert emergency responders when conversations with Meta AI suggest that a user—whether a teenager or an adult—may be at imminent risk of taking their own life.
It said the initiative builds on its existing suicide prevention efforts across Facebook and Instagram, where the company already notifies emergency services about credible suicide threats.
Meta disclosed that it made more than 19,000 referrals to emergency responders worldwide in the past year, leading to welfare checks on individuals considered to be at risk.
To improve the quality of its responses, the technology company said it consulted more than 75 clinicians specialising in adolescent mental health to review how Meta AI responds to prompts relating to suicide and self-harm.
The experts evaluated hundreds of AI interactions and provided recommendations on making responses more empathetic while directing teenagers to professional support and trusted adults.
Meta said it had incorporated the recommendations into its AI system to ensure more supportive and age-appropriate conversations.
Also Read: Listicle: 10 Essential Meta Safety Tools Every Parent and Teen Should Know
The company also announced that its stricter Limited Content setting for Instagram Teen Accounts would now apply to Meta AI conversations.
According to Meta, the setting further limits the types of conversations teenagers can have with the AI assistant by declining to respond to a broader range of sensitive prompts.
The parental alert feature is already available to supervising parents in the United States, the United Kingdom, Australia and Canada, while global rollout is expected before the end of the year.
