Direct transcription seems fine and all, but I find AI assessment of things like "Agitation, Defensiveness, Stress, Aggression" informing police work to be very troubling.

If the spam is better quality than the human written stuff, who's to say we aren't better off?

Quality in this case doesn't necessarily mean ground truth accuracy - it just means ability to look accurate to humans.

I agree, that's the problem, but I think it's still somewhat complicated.

Imagine someone posting an extremely well written and insightful postmortem of an outage. It would show advanced and accurate usage of all kinds of tools to get to the bottom of the outage. It would be extremely useful reading for anyone investigating a similar outage, but the outage never actually occurred.

Now you have both ground truth accuracy and misleading fiction at the same time. Whether or not that makes the post useful depends entirely on the conclusions you're drawing from it.