It has nothing to do with confirmation bias. It has to do with having accurate, real information. Knowing what the heck you're talking about.
First of all, objective truth is a real thing. A thing is either true or false. It doesn't matter how many millions agree or disagree with it.
Asking Google or secular scientists about Canon Law isn't "more accurate" just because they're clueless about Canon Law or Church History. Or because they're going against me, a "religious" person. A Traditional Catholics isn't always wrong just because he's a minority, or because "religious people are backwards".
Having an AI that has objectively true (trusted) information on these topics is very important. An AI is only as good as the information that goes into it. Either during training, after training, etc.
Do you not realize that "mainstream" sources don't always have the best information? Do you not think popular sites, sources can be wrong?