>While I understand the intent, it often makes the text verbose and awkward. So usually the suggested text is safer, but at the same time harder to read

this seems a side effect of being trained on the internet corpus. Much of online communication of the past decades has been people forcing themselves to write unnaturally in order that someone with an ax to grind does not willfully interpret something in the worst way possible in order to elevate their moral stature at the expense of those they interpret.

A somewhat dumbed down version of much of the latter half of 20th century literary criticism, in my opinion.

At any rate, the AI, as in every other thing it does, learned it from watching us.