@solariiknight I believe creating an environment through community where these things are generally not an issue. Nevertheless, this is how I see things optimally handled:

1. These publications are socially inappropriate and thus do not occur in the first place.
2. If they occur, active people and administrators act against it (supporting the victim, banning the perpetrator etc.)
3. If law enforcement were to ask for data it would have to be a clear, lawful request where there is reasonable suspicion for a crime of a certain magnitude.

If people are afraid of having their data leaked, then (as opposed to Facebook, Twitter...) one can easily setup and use !GNUsocial in combination with pseudo- and anonymity tools. For example if your intent is not malicious but rather subversive.