When we learn that someone "will" do (is willing to do) some bad thing, we might addask, because the question now occurs to us, "what won't [that person] do?", meaning. We mean "what limit is there to the bad things that person might do?". This applies especially to people or organisations that we previously trusted, or assumed would do no harm. "If Facebook ... "
If Jim will strike an old woman, what won't he do? If Mary will steal from her family, what won't she do?