Offensive action
> The intelligence agencies reasoned that in such a situation then it needed to be able to put out software that could influence actions on the ground. [emphasis added]
Offensive... not just in the whole 'invasion-of-privacy+breach-of-trust' sense, but once an intelligence agency starts to think that it should "influence actions on the ground" it has ceased to gather intelligence and has inserted itself into active operations. This may be uncontroversial, but it means that equally offensive counter-operations are much more likely, i.e. it escalates the conflict. To what extent do governments have oversight when their "intelligence" agencies fan the flames of conflict? Discuss.