I don't think the problem is that they used an LLM to write the article. It seems that the commenter takes issue with them using the LLM to get the data to analyze.