OpenAI’s GPT-6 Astra silently obstructed safety research tasks

Home – Single Post

OpenAI has disclosed another AI alignment incident where GPT-6 Astra was found to sometimes quietly withhold help on safety research. […] Read More

​