OpenAI Introduces New ‘Trusted Contact’ Safeguard for Cases of Possible Self-Harm
On Thursday, OpenAI unveiled a new feature called Trusted Contact designed to alert a trusted third party if mentions of self-harm are expressed within a conversation. This feature allows an adult ChatGPT user to designate another person as a trusted contact within their account, such as a friend or family member. In cases where a conversation may turn to self-harm, OpenAI now encourages the user to reach out to that contact.
OpenAI has faced significant legal challenges from families of individuals who have committed suicide after interacting with its chatbot. In several instances, these families claim that ChatGPT either encouraged their loved ones to kill themselves or assisted in planning such actions.
The company currently employs a combination of automation and human review to address potentially harmful incidents. Certain conversational triggers activate OpenAI’s system to identify suicidal thoughts, which then inform a human safety team. According to OpenAI, every time it receives such notification, the incident is reviewed by a human within one hour.
If OpenAI’s internal team determines that an incident represents a serious safety risk, ChatGPT will notify the designated trusted contact via email, text message, or in-app notification. The alert aims to be brief and encourages the contact to check in with the user. It does not include detailed information about what was being discussed, prioritizing privacy for the user.
The Trusted Contact feature complements the safeguards introduced by OpenAI last September that granted parents oversight of their teens’ accounts through safety notifications designed to alert them if OpenAI’s system perceives a “serious safety risk.” For some time now, ChatGPT has also included automated alerts prompting users to seek professional health services in case conversations lean toward self-harm.
Crucially, the Trusted Contact feature is optional and can be activated on any account. Even if protection is enabled on a particular account, users have the flexibility to maintain multiple ChatGPT accounts. OpenAI’s parental controls are also available as an option, sharing this limitation.
In its announcement post, OpenAI stated, “Trusted Contact is part of our broader effort to build AI systems that help people during difficult moments. We will continue to collaborate with clinicians, researchers, and policymakers to improve how AI systems respond when individuals may be experiencing distress.”
- The Trusted Contact feature complements existing safeguards for teen accounts.
- It allows users to designate a trusted contact in case of self-harm mentions.
- The feature is optional and can coexist with parental controls.
Key Takeaways
- OpenAI introduces Trusted Contact to alert a trusted third party in cases of self-harm mentions within conversations.
- The feature complements existing safeguards, including parental controls and automated alerts for professional health services.
- The feature is optional and can coexist with other safety measures offered by OpenAI.
Stay ahead of AI. Get the most important stories delivered to your inbox — no spam, no noise.
