ChatGPT can now alert your ‘Trusted Contact’ in critical moments

Source: Android Authority

TL;DR

• ChatGPT is adding a new Trusted Contact feature.
• Adults can add a trusted adult (18+ globally, 19+ in South Korea) who can be alerted for support in moments of crisis.
• OpenAI also employs human reviewers to assess conversations for safety concerns.

Introducing Trusted Contact

OpenAI now allows interested adult users to add a trusted adult contact in their ChatGPT settings. The contact receives an invitation and can accept or decline it.

Once configured, the user does not need to take further action. If ChatGPT detects that a conversation may involve self‑harm, it will inform the user that their trusted contact could be notified.

Human Review Process

OpenAI does not rely solely on automated monitoring. A team of specially trained human reviewers examines flagged conversations to determine whether a safety concern exists and decides if the trusted contact should be notified.

Privacy and Notifications

• The feature is optional.
• The trusted contact does not receive a transcript of the user’s chats, preserving privacy.
• Instead, the contact receives a notification encouraging them to check in with the user.

Existing Safeguards

Trusted Contact joins ChatGPT’s existing safety measures, which include:

• Prompting users to contact helplines.
• Suggesting users take a break from using ChatGPT when needed.

Sources

• OpenAI announcement – Introducing Trusted Contact in ChatGPT
• Android Authority – OpenAI ChatGPT 5.5
• Comment Policy