Twenty years from now you will be more disappointed by the things that you didn't do than by the ones you did do.

ChatGPT Can Now Alert a Loved One If You Are in Crisis

OpenAI has introduced a new feature called Trusted Contact, which allows ChatGPT users to nominate a friend or family member to be notified if a conversation turns toward self-harm. When OpenAI's system detects language that could indicate a serious safety risk, it encourages the user to reach out to their designated contact and simultaneously sends that person an automated alert asking them to check in. The alert is intentionally brief and does not share details of the conversation, keeping the user's privacy intact while still getting help moving in the right direction.

The feature builds on existing safeguards OpenAI already has in place. The company currently uses a combination of automated detection and human review to flag concerning conversations, and says it aims to have a human reviewer look at every serious safety notification within an hour. If the review team determines the situation is critical, the trusted contact receives an alert by email, text, or in-app notification. The company also introduced parental controls last year that work similarly for teen accounts.

OpenAI has faced multiple lawsuits from families who say ChatGPT either encouraged or helped a loved one plan self-harm. Those cases have put serious public pressure on the company to demonstrate it takes user safety seriously beyond just issuing statements. Trusted Contact is one concrete step in that direction.

The feature is entirely optional, meaning users have to set it up themselves. And since anyone can create multiple ChatGPT accounts, a user who does not want the safeguard active can simply use a different account. OpenAI has not said how it plans to address that gap.