> My only hope is that a benevolent AI will coordinate this, having prime directives not to harm humans ( by the means of interfering with the communication ).

Mmm...who's going to write that benevolent AI? :)