Aside from emergent behavior, are any of the items on that list unique to AI? They sure don’t seem it; they’re either broadly applicable to a number of already-available technologies, or to any entity in charge or providing advice or making decisions. I dare say even emergent behavior falls under this as well, since people can develop their own new motives that others don’t understand. Their advisory doesn’t seem to amount to much more than “bad people can do bad things”, except now “people” is “AI”.