I am catching up on my economics reading and was referred to this paper by Daniel Klein and Michael Clark from George Mason University. As I began reading, one section really resonated with me, especially if you replace the word "liberty" with "risk" and change the example to reflect information security:
Even when it is unambiguous that an action, considered in its direct aspect, is liberty reducing, it might be liberty augmenting considered in a larger aspect. For example, taxing people to wage war and dropping bombs on others are liberty reducing in their direct aspect, but if the war topples a Saddam Hussein, it might be liberty augmenting in its larger aspect. Thus, again, we have ambiguity about whether the action is liberty augmenting. This ambiguity arises not from ambiguity in any local facet of the action, but in “summing” over the facets. If all the facets go in one direction, either all reductions or all augmentations, there is no such ambiguity. But when some facets are reductions and some are augmentations,then it might be very difficulty, even impossible, to assess the action in terms of liberty. The difficulty stems from two problems: First, weighing the set of pluses against the set of minuses; second, knowing what is in each set. Saddam Hussein was a brutish fellow leading a highly coercive government, but do we know that the wider facets of toppling him were pluses for liberty?