Because AI is famously not at all biased in its fact checking.

Surely having a dumb mathematical model that averages all trained input to censor anything it considers untrue based on that average is a great way to keep people in line with the status quo achieve freedom on the internet.