You'd basically need an AGI to actually evaluate how biased an article is.

Too many bazingas, including a few on Hexbear, believe that a sufficiently large dataset (and sufficient burned forests and dried up lakes) will make the treat printers AGI.