Content platforms have no place policing content. All the algorithms currently used are utterly flawed, and they will always be flawed, both technically and morally. Technically, they'll always have false positives, especially at scale. Morally, they turn the platform owners into political figures inherently, siding with or against certain thoughts and sentiments.
The only solution is for platforms to be completely content agnostic and allow absolutely everything except outright graphic violence and adult content.
No removing any kind of speech whatsoever, including so-called "hate speech", no policing political topics, no identifying "fake news" and so forth -- words cannot hurt anyone, objectively, and everyone has their own personal agency to decide what they want to watch and make their own decisions based on what they see, even if it's state-funded propaganda.
If the police or FBI come to them with a warrant and take down notice, then sure. Otherwise, allow absolutely everything, and allow the free market of popularity to reign supreme.
In fact, all sides of all political and social issues should be outraged that content platforms think they have a place deciding what we can and cannot see.
And advertisers should stop allowing online outrage to dictate where they advertise by realizing two simple truths: a) if I'm watching a video with your advertisement on it, it's because I made the conscious choice to watch that video, and even if the video is about apartheid or waterboarding or guns, it's content that I want to consume, so they have no place stripping me of my personal agency, and b) the old model of advertisers being seen as endorsers for TV shows doesn't apply to on-demand online content.