removing state propaganda bots/trolls from an instance is a job for the instance moderators.
they must have an understanding of what propaganda and fake news look like, so that they can handle this content when they encounter it.
the approach used by counter.social of banning a bunch of "russian" IP blocks and proxy lists will not stop propaganda from being distributed via their service, because the fake news was being relayed by hacked machines in the US.
bottom line: skilled moderators are needed to combat fake news, there is no easy technical solution.