If you want to preserve classification, I'd work on the training and motive side, rather than what happens after a leak. Prevent the leaks by preventing activity that motivates leakers. Punish anyone who sells information for direct cash or commercial favors. Make it clear what information will result in immediate physical harm if leaked, and don't abuse the classification system to hide embarrassing, negligent, corrupt, or criminal behavior. Don't engage in mass surveillance.

In general when I solve a problem I try to go as close to the source as possible, and/or look at a layer of abstraction that isn't often considered.