OpenAI launches Trusted Contact to alert friends of mental health concerns
openai launched 'trusted contact,' a safety feature for adult chatgpt users. it allows users to designate a person to receive a notification if the system detects self-harm or suicide topics. notifications do not include chat transcripts and are reviewed by a specialized human team before sending.
OpenAI introduced 'Trusted Contact,' an opt-in safety feature enabling adult users to designate a person to receive notifications if the AI detects potential self-harm or suicide discussions. The system expands on existing parental controls introduced in September, responding to previous incidents involving minors and AI interactions. Notifications are deliberately limited in scope—they omit chat transcripts—but are reviewed by a 'small team of specially trained people' before alerts are sent via email, SMS, or in-app message.