ChatGPT Trusted Contact Feature Will Notify Someone If It Detects Self-Harm Concerns

OpenAI is introducing a new safety feature in ChatGPT that could help users during emotional or mental health crises. The feature, called Trusted Contact, allows adult users to choose one person who may be notified if ChatGPT detects signs of serious self-harm risk during conversations.

As AI chatbots become more common in everyday life, many people are using them to discuss deeply personal topics. Some users turn to chatbots for emotional support, advice, or simply someone to talk to during difficult times. While AI can provide information and basic guidance, it cannot replace real human care or professional mental health support. OpenAI’s new feature appears to recognize that limitation and adds a layer of human connection when needed.

ChatGPT Trusted Contact Feature Will Notify Someone If It Detects Self-Harm Concerns

The Trusted Contact feature is currently rolling out to adult users through ChatGPT’s settings menu. Setting it up is completely optional. Users can choose one trusted person, but that person must meet the minimum age requirement. In most countries, the contact must be at least 18 years old, while users in South Korea must choose someone who is at least 19.

Once selected, the trusted person receives an invitation explaining what the role involves. They have one week to accept the request before the feature becomes active. If they decline, the user can select another person instead.

One important detail is that the alert system is not fully automatic. If ChatGPT’s systems detect a conversation that may involve self-harm concerns, the chatbot first warns the user that their Trusted Contact could be notified. It also encourages the user to reach out directly by offering suggested conversation starters or ways to begin the discussion.

After that, a trained human review team examines the flagged conversation. OpenAI says only a small group of specially trained reviewers can access these situations. If the team confirms there is a serious risk, the Trusted Contact receives a notification through email, text message, or an in-app alert.

The company has also stated that privacy remains a major priority. The alert sent to the trusted person does not include chat transcripts, screenshots, or detailed conversation content. Instead, it simply explains that the user may be going through a difficult situation and encourages the contact to check in with them.

According to OpenAI, the company aims to complete the human review process within one hour. This is designed to make the feature more responsive during potentially urgent situations.

See Also: ChatGPT Gets Smarter with New GPT-5.5 Instant Default Model

Trusted Contact is also part of a broader effort by OpenAI to improve safety on its platform. The company had previously introduced safety alerts for linked teen accounts, allowing parents to receive warnings if signs of emotional distress appeared. The new feature expands that idea for adult users.

OpenAI reportedly worked with mental health experts, researchers, and organizations like the American Psychological Association while developing the feature. Even so, the company clearly states that ChatGPT is not a replacement for crisis hotlines, emergency services, therapists, or medical professionals.

The introduction of Trusted Contact highlights a growing reality: people are increasingly having emotional and personal conversations with AI systems. By adding a way to involve trusted humans during serious moments, OpenAI is trying to create a safer balance between technology and real-world support.