Is Apple’s AI adding hallucinated details? The last I read it’s just used to merge multiple images - up to 8 or 9 images - to form the final
image. While I could see details getting lost or artifacts being added, I don’t think it can add actual “feature” details that don’t exist.
It already does some amount of features detection and targetted enhancing. Things like making this face smoother, or this sky bluer, or this grass greener, etc. From there, I wouldn't be too surprised if some details get added beyond what was strictly captured (e.g. bubbles in drinks, leaves in trees, …)
