OpenAI's Trusted Contact Turns ChatGPT Into a Mental Health Safety Net

6. OpenAI's Trusted Contact Turns ChatGPT Into a Mental Health Safety Net

OpenAI is rolling out an optional feature called Trusted Contact for ChatGPT, announced May 5, 2026. Adult users can designate a friend, family member, or caregiver to receive automated notifications if OpenAI's systems detect that a conversation has touched on self-harm or suicide. The feature is opt-in, applies only to adult accounts, and positions OpenAI explicitly in the mental health safety space for the first time at a product level.

This move reframes the competitive pressure on every major consumer AI assistant. Google's Gemini, Anthropic's Claude, and Meta AI all serve tens of millions of users who discuss personal struggles with chatbots daily, but none have shipped a comparable crisis-notification layer. By building this infrastructure now, OpenAI sets an implicit industry standard before regulators do. The EU AI Act's high-risk classification system already flags emotional support applications, and U.S. congressional scrutiny of AI's role in youth mental health has intensified through 2025 and into 2026. A proactive safety feature gives OpenAI a defensible posture in both arenas. It also deepens user lock-in: once a contact network is established inside ChatGPT, switching costs rise in ways that have nothing to do with model quality.

The broader pattern here is OpenAI moving ChatGPT from a general-purpose tool toward a platform with social infrastructure. Trusted Contact follows the memory feature, custom instructions, and operator controls in building a layer of personalization that compounds over time. Watch whether Anthropic responds with something similar for Claude, and whether Apple or Google surface equivalent features at the OS level before the next major platform release cycle. The company that owns the crisis-notification layer owns a category.

Source: ChatGPT's 'Trusted Contact' will alert loved ones of safety concerns