How so? Could they not ML-detect the main body/content of the post and do a simple diff to previous "crawls" of the same content? You would think this gets picked up and somehow "resolved" because it's essentially a duplicate post or entry in their crawled pages DB.