The answer to this is that according to the reality today, even in Delma matters, it is impossible to trust the artificial intelligence without checking if the things the software says are accurate, and it is known that even the sources that the artificial intelligence cites are not correct in many cases, and this is still before we come to discuss the "decision" or "judgment" of The artificial intelligence (without citation) that it should not be said that it cannot be trusted, and even if in the future they develop new software that will only quote correct things, there will still be a need for a halachic decision by a sage for practical halacha.
And according to what I commented in another answer, there is a general problem with talking to artificial intelligence about halachic matters as long as the software involves opinions that are contrary to halachic law.
And this is apart from what the arbiters of the generation have already warned against using this software.
מק"ט התשובה הוא: 4055 והקישור הישיר של התשובה הוא: shchiche.com/4055
