It's a mostly human process, and the humans that pay money are the ones most likely to be heard first. Many social media systems experimented with a purely aut

It's a mostly human process, and the humans that pay money are the ones most likely to be heard first. Many social media systems experimented with a purely automated system and had difficulty with automated systems that likewise reported everything that they didn't like and that resulted in the content creators getting banned for non-reasons. This leads to needing to having a human check things - and humans don't scale. Look at the stories of Youtube content moderation - both the humans involved and the false positives from when humans aren't involved - for examples of "why we can't have nice things."