Our paper contains a small bit of application of graph theory to something related to biophysics. A referee claims that we should mention other applications of graph theory to biophysics in the introduction. The "keywords" he gave us are actual titles of papers, and all but one has one common author. They are also not really relevant (which we pointed out), not really the important ones in that field (there are other, earlier papers doing very similar things), and this we have pointed out. He still insists, that according to him "as an expert", more references should be added. Now, I checked the guy who wrote those papers, and he had 60 papers in the last three years, is a professor somewhere in the middle east, and has no publications before those. Is there some way to check if the papers he want us to cite do not simply come from a paper mill? (Note, that this is a theoretical topic. I cannot check, e.g., measurements for having the same noise.)
I don't think it's up to you to show that the papers are from a paper mill or anything else to a standard that would convince a court beyond reasonable doubt.
Like Jon Custer states in a comment, a competent editor will understand this request is problematic and, like you, will find it worrying that these papers all share this suspicious author.
I'd write a note to the editor, separate from a response to the reviewers, raising your concern. It probably would have been best to do this in the first place but you can still do it now.
It's up to the editor, not this reviewer, to decide the fate of your paper. If the editor is also insisting on these additions, I'd question the editorial quality of the journal and likely resubmit elsewhere. If you're absolutely convinced of the quality of this journal, then it may be worth escalating above the editor.