Just technical problems in their “banned topic” identification models. No need to be concerned.

The point is that people are more aware of problems happening with that topic, but ignore whether it also happens with other topics. So at the moment it's a very skewed view.