The Observer investigates OpenAI's safety unit
On Wednesday, we covered a family suing OpenAI after their daughter's death. Now a bigger investigation looks at the system behind such cases. The Observer, a UK newspaper, published a report on OpenAI's safety unit. Its title quotes a family member: 'If it wasn't for ChatGPT my brother would still be alive.'
The Observer says OpenAI's race for dominance has compromised user safety. It links conversational AI to psychiatric crises and violence. The story was shared on r/artificial. Our source only gives the summary, so we do not have the details of each case or OpenAI's answer.
Expect more pressure on OpenAI to show what its safety team really does, and with what resources.
Many people talk to chatbots at their most fragile moments. If speed wins over safety inside the leading lab, those users pay the price. The real question: who checks the people who are supposed to check?
