Also that such filters also double trips over itself when people write in other languages... if they just matches words or even just parts of words.if its reported, yes. Auto-censorship is so strict it filters out a near-endless stream of phrases which they cannot manually check or punish.
add to this creative alternative spellings like l33t sp3@k and other ways to convey UR msg. Lets see what horrors we might get with the future of AI chat moderation...