How do you know if you can trust then if they are human generated? Your trust the people. AI isn't going to jump and just generate docs, a person has to prompt it, and you should expect the person to proof read and correct it. If the person turns out to be untrustworthy, you stop trusting them.