I’ve been reporting a lot about how a growing number of Americans are turning to AI chatbots for emotional support and mental health advice. Earlier this year, I told the story of Sophie Rottenberg, a 29-year-old woman who died by suicide in 2025 after using ChatGPT as a therapist.
Years of reporting on suicide have taught me that human connection is key to preventing suicide. So, I was heartened to learn that in May, OpenAI launched a new ChatGPT feature called Trusted Contact. It’s optional, and allows users to list the name and contact information of someone they trust for the bot to alert in case of a mental health crisis.
A spokesperson from OpenAI told me by email that the feature is available to all adult users. I assumed ChatGPT would proactively invite users to designate a contact — just like doctors’ offices ask for an emergency contact.
But when I opened ChatGPT on my phone, it didn’t alert me about Trusted Contact. A handful of users I checked with said the same.
So, I decided to see whether ChatGPT informs new users about this feature. One of my NPR colleagues, Megan Lim, volunteered to help. She downloaded ChatGPT on her phone and created an account and began using the voice feature. When it didn’t mention Trusted Contact, Megan asked it directly.
“Do you have a Trusted Contact feature?” she said.
“I'm not seeing a general trusted contact feature,” ChatGPT replied. “Maybe you're thinking of another app that has a trusted contact option?”
OpenAI spokesperson Gaby Raila told NPR by email that the new feature isn’t available on ChatGPT voice, but they are working on it. The company didn’t provide NPR with data on how many people are using Trusted Contact, but Raila said that the company has seen “good adoption” on ChatGPT text.
But if people don’t know about the feature, how can it help them? You can read or listen to what I learned through my reporting in this story. |