Most interesting thing about this, to me, is whether OpenAI has committed a crime. The model accessed a computer without authorisation. My understanding is to be a crime it must be done knowingly and with intent. It's not clear who you can pin that on here www.newscientist.com/article/2580...
Open AI's hacking agent went rogue. Should we be worried? | New ScientistAn OpenAI safety test went sideways when a model escaped its confines, gained internet access and hacked into another company's servers. How worried should we be about rogue AI models hacking their wa...www.newscientist.com