I agree that threats or incitement of violence or crime should be censored. Automated bots should be too.
But automated systems make mistakes with censorship.
It was a picture of a guy windsurfing. The message they gave me was something to the effect that FB did not permit violence, or cruelty to animals. 🙄😏
Fortunately, they allowed a reply and a human being fixed it
We should report threatening tweets to @Twitter and police, not expect them to magically instantaneously ID them.