OpenAI has introduced a new “Trusted Person” feature for ChatGPT, which will send warnings to the user’s loved ones if mentions of self-harm appear during a conversation with the chatbot. This is reported by TechCrunch with reference to the OpenAI press service.
The new feature will allow an adult ChatGPT user to add a trusted contact to their account — for example, a friend or relative.
If the user writes about suicidal thoughts during a conversation, OpenAI will suggest that they contact this person. At the same time, the chat will send a message to a trusted contact asking them to reach out to the user.
Such notifications will come via email, in-app, or text message. According to the company, the message will be brief and should encourage the contact to reach out to the user. It will not include details of the conversation to maintain privacy.
OpenAI has faced a wave of lawsuits from families of people who committed suicide after talking to its chatbot. In a number of cases, families claim that ChatGPT pushed their loved ones to commit suicide — or even helped them plan it.
Previous research found that OpenAI’s free chatbot doesn’t always detect risky behavior or challenge false beliefs in users with mental disorders.
