> This would allow them to access millions of user conversations that are unrelated to the case

What I don't understand is why they can't have a third party handle the data. Why does the NYT need it itself?