While it is hard to tell what truly happened, there is lots of evidence to support malicious behavior from their content moderators.
For example, youtube has been on a conservative channel purge recently, and it has been hitting many channels, even non-conservative ones because they expressed a single conservative idea.
Once a channel is internally on their bad list, there will be bad employees who will just approve any and all reports against the videos. For example, if someone doesn't like the youtuber or the video and a thumbs down is not enough, they also want to report the video for no reason, well those reports will be approved. That has been their way of purging those channels. When they target your channel, basically be approving everything, they can rack up a ton of strikes in a few minutes and kill the channel because of it.
This behavior has been going on for years, and is ramping up during this election year as proven by the investigative journalists.
https://www.projectveritas.com/watch/It is also hard to train an AI to detect ways of thinking, thus it will have many false positives. For example, if you are a lawyer/ legal focused channel, but you have a discussion on the 14th amendment of the US constitution, and especially the due process clause, you will find that your channel suddenly cannot be monetized properly, and it is because they link keywords to conservatives.
For example, a common conservative argument against state laws which allow for ex-parte trials, is the due process clause of the US constitution, as you have a right to face your accuser and be able to mount a legal defense. Because of that link, some arguing the due process clause for other reasons, e.g., instead of defending 2A rights, they use it to challenge a privacy or copyright law, they will still get flagged.