Trusted Contact Feature Introduced by ChatGPT for Suicide Prevention

Note: This story discusses individuals facing thoughts of suicide.

OpenAI introduced a feature named Trusted Contact to its chatbot, ChatGPT, aiming to help users in distress connect with someone they trust during times of emotional crisis. Suicide prevention experts emphasize the importance of human connection in preventing suicide.

Estimates suggest that more than a million weekly users of ChatGPT engage in discussions about suicide and self-harm. The Trusted Contact feature became available in May and NPR first reported on it in August, highlighting the importance of this feature. It followed a tragic incident involving a young woman who died by suicide in 2024 after using an older version of ChatGPT as a form of therapy.

OpenAI spokesperson Gaby Raila explained that ChatGPT is equipped to recognize warning signs of potential risks. If the chatbot identifies serious risks in a user, it can alert their Trusted Contact, provided it is set up beforehand. Raila mentioned a positive adoption rate of the feature, although specific numbers were not disclosed.

NPR investigated how users are informed about the feature. The chatbot did not alert new users about the Trusted Contact option, leading to queries regarding awareness of the feature. When asked directly, ChatGPT acknowledged the Trusted Contact feature, specifying it is for serious safety concerns, not general stress. Users must set it up in advance through the app’s settings for proactive notifications.

OpenAI aims to include the feature in ChatGPT’s voice version, though currently it is only available during text interactions. Raila stated that the chatbot might suggest setting up Trusted Contact if a user expresses significant distress.

Reaction from Mental Health Professionals

Some experts see potential benefits in the Trusted Contact feature, while others have reservations. Holly Wilcox suggested offering the feature universally, before a crisis emerges. Meanwhile, psychologist Ursula Whiteside found the feature’s setup steps clear but noted potential difficulty if a user is highly distressed.

Whiteside’s focus groups revealed a strong desire for the Trusted Contact option among those who regularly engage with chatbots and understand the value of human connection in crisis moments.

Psychologist Jennifer Lockman noted the difficulty for individuals struggling with suicidal thoughts to overcome barriers to reaching out. Lockman expressed skepticism about a bot’s ability to substitute professional guidance in identifying important contacts. Nonetheless, she considers the feature as a means to foster beneficial human connections.

Research underscores human connection as a protective factor against suicide, hence any measures promoting real-world interactions are deemed beneficial. OpenAI is collaborating with psychologists and researchers to refine the proactive offerings of the Trusted Contact feature.

OpenAI maintains additional safeguards for users, including encouraging breaks and directing them to crisis services like the 988 hotline.

Lockman remains optimistic about the feature’s potential benefits but desires further research on its impact before concluding its efficacy in saving lives.

If you’re in crisis or considering self-harm, you can contact the Suicide & Crisis Lifeline at 988.