OpenAI is introducing a raft of new safety measures after being sued by the parents of a 16-year-old who alleged ChatGPT acted as his “suicide coach.” The company said it plans to introduce new features, including curfews and regulating what it remembers from past chats. It will also allow control over whether chats are used for training its models and prevent access to sensitive content. It comes after OpenAI and its CEO, Sam Altman, were sued by the family of Adam Raine for wrongful death earlier this year. In the filing in California Superior Court, his parents, Matt and Maria, claimed, “ChatGPT actively helped Adam explore suicide methods.” The new system will require both the child and the parents to accept an invitation to link their accounts. Parents will also be able to limit image generation and voice responses. The new features won’t give parents access to transcripts, although they may receive alerts if moderators or the program detect certain indicators, Reuters reports. If you or a loved one is struggling with suicidal thoughts, please reach out to the National Suicide Prevention Lifeline by dialing or texting 988.
Read it at Reuters
