Cheryl has made a fine point. I, nor other safety professionals, have really said much about general safety precepts as they relate to AI, as opposed to the bullshit artist millenarian cultists that are happy to do so at length. One of the reasons is "These tools are an active detriment to safety."
There are general safety precepts that can be applied to AI as much as to handling plutonium. I've been kind of wondering why @funranium hasn't said much about this. The details are outside my experience, but the precepts seem to be unknown to those doing the experiments.