They didn't like having to openly apply human judgement so they applied AI. AI that has some human judgement built in. And bias along with it. But now nobody is responsible because the machine did it.
TFA appears to address this, the automated system is used to help identify comments for human moderators. The human is the one responsible. And according to stack overflow they plan to use this system only to help the human moderators.
If you don't believe them and suspect they will have a fully AI flagging system in the future (in direct contradiction of this post) thats fine but it might be better stated as such, since human judgement is currently still paramount.
If you don't believe them and suspect they will have a fully AI flagging system in the future (in direct contradiction of this post) thats fine but it might be better stated as such, since human judgement is currently still paramount.