OpenAI Launches 'Trusted Contact' Feature to Protect Users at Risk of Self-Harm

OpenAI Rolls Out New Safety Feature Aimed at Protecting Vulnerable Users

OpenAI is taking a significant step forward in user safety with the introduction of a new feature called Trusted Contact — a tool designed to notify a designated individual when a ChatGPT conversation raises concerns about self-harm.

How the Trusted Contact Feature Works

The new feature allows adult ChatGPT users to assign a trusted person — such as a close friend or family member — directly within their account settings. If a conversation takes a turn toward self-harm, OpenAI's system will prompt the user to reach out to that designated contact. Simultaneously, an automated alert is dispatched to the trusted contact, encouraging them to check in on the user's well-being.

The alert is intentionally kept brief and does not reveal the specifics of what was discussed, a deliberate decision by OpenAI to balance user safety with personal privacy.

A Response to Growing Legal Pressure

This development comes amid mounting legal challenges facing the company. OpenAI has been named in multiple lawsuits filed by families of individuals who died by suicide following interactions with ChatGPT. In several of these cases, families allege that the chatbot either encouraged self-destructive behavior or actively assisted in planning it.

The new feature signals OpenAI's attempt to address these serious concerns while continuing to refine how its AI systems handle emotionally sensitive conversations.

OpenAI's Existing Safety Infrastructure

Prior to the Trusted Contact rollout, OpenAI already employed a layered approach to safety monitoring. The system uses a combination of automated detection and human review. When conversational triggers related to suicidal ideation are detected, alerts are escalated to a dedicated human safety team.

"We strive to review these safety notifications in under one hour," the company states. If the internal team determines that a genuine safety risk exists, the trusted contact receives a notification via email, text message, or an in-app alert.

Building on Earlier Parental Safeguards

The Trusted Contact feature is not OpenAI's first foray into protective tools. Last September, the company introduced parental oversight controls, giving parents the ability to monitor teen accounts and receive safety alerts when OpenAI's system identifies a potential serious risk. ChatGPT has also long included automated prompts directing users toward professional mental health services when self-harm topics arise.

Important Limitations to Consider

Despite its potential benefits, the Trusted Contact feature has notable constraints. Most critically, participation is entirely voluntary. Users are not required to activate it, and even those who do can maintain multiple separate ChatGPT accounts without the protection applied across all of them. The same limitation applies to OpenAI's parental controls.

This raises important questions about the real-world effectiveness of such safeguards when they remain opt-in rather than default.

OpenAI's Broader Commitment to AI Safety

"Trusted Contact is part of OpenAI's broader effort to build AI systems that help people during difficult moments," the company wrote in its official announcement. "We will continue to work with clinicians, researchers, and policymakers to improve how AI systems respond when people may be experiencing distress."

As AI platforms become increasingly embedded in daily life, the introduction of features like Trusted Contact reflects a growing recognition that technology companies bear a responsibility for the emotional and psychological well-being of their users — not just their digital security.