In the case of publications we luckily have such fantastic resources as mediabiasfactcheck[0] to keep their bias in check and to keep them factual.
LLMs are much harder to fact-check because they can make anything up based on their training data and weights without sources.