The third option is for organizations to store loads of credible misinformation in addition to the "real" info. Probably this false information will be generated by software, so generating it will be very cheap. Hackers will have the problem of knowing what's real and what only looks real. Wikileaks will have difficulty maintaining credibility as hackers post lots of false data.
(It's not like this problem hasn't been solved by the intelligence agencies already.)
> Hackers will have the problem of knowing what's real and what only looks real.
So will legitimate internal members of the organization.
Haven't you ever had to wade through your company's intranet wiki trying to figure out which documents were are actually useful versus the giant piles of things that are outdated, wrong, superceded, etc?
That won't work because you'll be adding noise to your own data. You'll have trouble searching through your own data-sets for "real" stuff. If you do something to designate what's real, you're just back to square one.
(It's not like this problem hasn't been solved by the intelligence agencies already.)