This already is a thing people did at least as far back as I started getting into web privacy, which was ~10 years ago. I have been the target of it before.

LLM's are probably better at it, but I don't know if this is as destructive as people may guess it would be. Probably highly person dependent.

The micro-signals this paper discusses are more difficult to fake.

stylometry is only one aspect of de-anonymization. what you describe is certainly a threat that we will have to deal with, but there is a lot more to credible impersonation than just being able to mimic a writing style

So this means deanonymization doesn't work? Rejoice?

How to conduct a psy-op

https://youtu.be/YTGQXVmrc6g