How can you be sure what you're seeing is AI slop, as opposed to, for example, corporate-speak or just documents modeled on their past work wherever, etc?
The question on my mind: what do we do in a few years (or sooner) when the AI's have become good enough to pass the Turing test and we just can't tell anymore whether the response is AI slop or not?