As a stable diffusion user, I (subjectively) feel particularly adept in identifying AI generated images.
And its not just about spotting artifacts. There are specific subjects/styles SD models are really bad at reliably generating, or certain compositions it tends to like.