Journalists both have an incentive to generate a story, and potentially harm a competitor in the ad selling space. So they find as controversial of a video the

Journalists both have an incentive to generate a story, and potentially harm a competitor in the ad selling space. So they find as controversial of a video they can find (possibly with really low view counts), start a screen recorder, then hit refresh until a major brand's ad shows up before the video. Now they can run a headline like "Nike financially supports pro-genocide content on YouTube". It's not an entirely unfair thing either. That is something that is happening, just in miniscule amounts. So YouTube introduced brand safety options which predict how safe a video is and avoids such potential headlines. However since it's just an algorithm, it doesn't understand nuance. So now creators who are making educational content about tough subjects earn less money because they are bucketed as unsafe. Since unsafe content has less advertising biders, it earns less money.