Meta Deploys AI to Detect Teen Suicide Chats, Alert Parents

Meta has introduced a new safety feature that alerts parents when its artificial intelligence chatbot detects that a teenager may be discussing suicide or self-harm, as technology companies face increasing pressure to strengthen safeguards around young users’ interactions with generative AI.

The feature which is announced today is designed to identify conversations in which teenagers make clear references to harming themselves.

According to Meta, a dedicated AI safety system analyses chats for signs of potential crisis before the conversation is reviewed by trained personnel. Where the risk is confirmed, parents linked to the teenager’s account receive a notification encouraging them to check on their child and access mental health resources.

The development comes amid growing global scrutiny of AI chatbots following concerns from regulators, parents and child safety advocates over how conversational AI responds to users experiencing mental health crises, particularly teenagers.

Meta, explaining the new measure, said, “We’ve built a dedicated AI system that identifies conversations where a teen makes a clear reference to hurting themselves.” The company noted that such conversations will undergo manual review before any parent notification is sent to minimise false alerts.

Under the new system, parents will not receive the full content of their teenager’s conversations. Instead, they will be informed that Meta believes their child may need support and will be directed to guidance on how to have a constructive conversation about mental health.

The initiative builds on the social messaging company’s broader safety framework for Teen Accounts across Instagram, Facebook and other platforms, where the company has introduced stricter privacy settings, parental supervision tools and enhanced protections against harmful content for younger users.

The latest move reflects mounting pressure on AI developers to demonstrate that their systems can recognise and respond appropriately to vulnerable users. The rapid adoption of generative AI has intensified debate over whether chatbots should merely provide information or actively intervene when conversations suggest a user may be in danger.

However, the firm maintained that while its AI can detect warning signs, the technology is not intended to replace professional mental health services. Instead, the company said the feature is designed to encourage timely intervention by trusted adults by directing users to appropriate crisis support resources.

The announcement comes as regulators in several countries tighten oversight of digital platforms over child online safety. For instance, the United Kingdom has launched investigations into whether major technology companies are doing enough to shield minors from harmful online content under the country’s Online Safety Act.

For the social media firm, the new safeguard represents one of its most direct interventions in AI-assisted conversations involving teenagers, underscoring the growing expectation that technology firms must balance rapid AI innovation with stronger protections for vulnerable users.

Meanwhile, the rollout is expected to begin across Meta’s supported platforms in phases, with the company indicating that it will continue refining the system as it gathers feedback from parents, safety experts and mental health organisations.