In a proactive move to enhance user safety, OpenAI has launched a new feature for ChatGPT called "Trusted Contact." This initiative comes in response to growing concerns regarding the mental health implications of AI interactions. While experts advise caution in relying solely on chatbots for mental health support, many individuals are increasingly seeking assistance from AI platforms like ChatGPT.
The "Trusted Contact" feature allows users to designate a trusted individual in their lives who will be notified if the AI detects concerning discussions related to self-harm. This approach is not about sharing chat histories or collaborative interactions; instead, it aims to ensure that users have a support system in place should their conversations take a troubling turn.
Understanding How Trusted Contact Works
To activate this feature, users must select a contact who is at least 18 years old (19 in South Korea). Once selected, ChatGPT will send an invitation to this person to become a Trusted Contact. They have one week to accept the invitation, although they can decline if they prefer not to participate.
If the chosen contact agrees, the system becomes operational. Should ChatGPT identify discussions that suggest a serious safety concern regarding self-harm, it will notify the user about the possibility of reaching out to their Trusted Contact. Additionally, it will provide "conversation starters" to facilitate communication.
OpenAI emphasizes that the process is not entirely automated. A dedicated team of trained professionals will assess the situation. If deemed serious, ChatGPT will alert the Trusted Contact via email, text, or in-app notification, ensuring that the message contains only essential information without disclosing chat transcripts, thereby respecting user privacy.
The organization aims to review safety notifications within an hour and has developed this feature with insights from mental health professionals, researchers, and suicide prevention organizations. Participation in this program is entirely voluntary, allowing users to opt-in if they feel it could provide valuable support during difficult times.
This innovative approach could significantly enhance the way individuals connect with their support networks, fostering a more responsive environment for those grappling with mental health challenges.