Unless openai release the logs we have only their word that this was done fully autonomously and without their knowledge by an agent running their newest super

Unless openai release the logs we have only their word that this was done fully autonomously and without their knowledge by an agent running their newest super powerful model. For all we know they could have bought zero days and left them lying around for the agent to find. That may be unlikely, but it sounds less far fetched than an agent running a sophisticated attack against multiple targets over the course of four days and completely unbeknown to anyone at openai, despite the fact that they knew they where running a dangerous model with all safeguards disabled. So far there has been no comment about how the agent evaded monitoring and detection.