Meta will begin notifying parents if their teen appears to be discussing suicide or self-harm with Meta AI on Instagram, Facebook and Messenger, marking one of the platform’s biggest expansions of AI safety measures for young users.
The new feature, announced overnight, is rolling out immediately to parents using Instagram’s parental supervision tools in Australia, the US, UK and Canada, before expanding globally by the end of 2026. In Australia, children under 16 are not permitted to have an Instagram account under new online safety rules.
Meta is also actively looking into new ways for parents to manage how their teens use WhatsApp, and will be rolling out the features in the coming months. For younger users of parent-managed accounts on WhatsApp, there is no Meta AI available.
There have been several high profile cases of teenagers killing themselves after reportedly using social media apps like Instagram in recent years. In the UK, Molly Russell took her own life at the age of 14 in 2017, while a Malaysian teenager also reportedly killed herself after posting a poll on her Instagram account asking followers if she should die or not, and 69 per cent of responders voting that she should.
Until now, Meta AI has directed teens expressing thoughts of suicide or self-harm towards crisis helplines and encouraged them to speak with a trusted adult. Under the new system, parents will also receive an alert if Meta determines their teen’s AI conversation suggests they may be at risk.
Meta said the alerts are powered by a dedicated AI system developed with input from experts and parents, but stressed that every flagged conversation will be manually reviewed before a notification is sent.
If a teen’s intent is unclear, Meta said it will “err on the side of caution” and notify parents anyway.
The alerts will include expert resources designed to help parents navigate conversations with their teen about suicide and self-harm.
The announcement comes as AI companies face growing scrutiny over how their chatbots respond to vulnerable users, particularly teenagers.
Meta also revealed it is “building the ability to contact emergency services” if a conversation with Meta AI suggests that a user—whether an adult or teen—”may be at imminent risk of taking their own life”.
The company said the move builds on its existing suicide prevention work across Facebook and Instagram, where it alerts first responders to credible risks. According to Meta, it made more than 19,000 referrals to emergency services worldwide last year, helping first responders carry out welfare checks on people who may have been at risk.
To strengthen Meta AI’s responses to teens discussing suicide and self-harm, the company said it sought feedback from more than 75 mental health clinicians specialising in adolescent wellbeing. The clinicians reviewed hundreds of AI responses, with Meta saying the feedback is being used to improve how the chatbot handles these sensitive conversations.
Meta said this includes ensuring “Meta AI acknowledges the teens’ feelings when directing them to support resources, and doesn’t shut the conversation down too abruptly”.
The company also announced it is extending Instagram’s Limited Content setting to Meta AI chats. According to Meta, when parents enable the stricter setting, it will “further limit the kinds of conversations teens can have with Meta AI,” with the chatbot declining to respond to a broader range of sensitive prompts to reduce the likelihood of inappropriate interactions.
