“YouTube doesn't want you to have to see my 'disturbing' disabled body”

Oh absolutely, it rapidly falls apart when people employed by Google make choices about any of this. But Google is famously insistent on how humans don’t scale and for trying to automate everything with all their ai and algorithms. It feels like a real possibility that this entire process was only “supervised” by humans without any of them having an “actionable level of culpability” or whatever wording of choice the Google lawyers use to say “look people are involved but not involved enough you can sue us over this”. The level of stupidity we see with the ContentID system and it’s insane copyright claims on things like silence, white noises static, birds chirping, cats meowing, etc… it’s very very easy to imagine that an entire pathway of their flagging and moderation system exists to throw away false reports on obviously ok content or automatically accept flagging on content they have trained machine learning models on to cross check borderline cases, or even automatically flag and demonetise content matching previously flagged and moderated content even without a human ever seeing it. The process may be “match past flagged content -> auto flag -> user requests review -> AI checks it against a model trained on other data -> still rejected -> user re-requests -> Google algorithms check some other thing before deciding if it’s even worth showing to a real person -> final rejection without even showing it to a person.

All this kind of stuff is private info Google doesn’t share and doesn’t want to share about internal processes and would have to be compelled legal disclosure involving an entire round of “sub-lawsuits”, blocking meaningful progress of the main lawsuit before it can even be established just how culpable Google is either by a specific human employee deciding, or some complex web of obligation on them being required to design to fulfill legal obligations such as not discriminating against protected classes.