Or any other issue.
It will only do what it's programmers deemed acceptable.
No matter what they try to tell you it is not AI. Just another data retrieval system.
I did some further digging into how ChatGPT was trained. Apparently it is shown tons of articles as its training data.
I would tentatively say, based on that dialog, that the folks who selected the training data were very careful to include only acceptable ruling class narrative articles. That the bot claims to have NO information about Mayorkas lie on whippings means someone had to curate the training articles pretty carefully.