But in any case, the difference is huge between writing a one-sentence prompt and choosing and arranging a motive, arranging the photo setting & lighting (unless it's a Paparazzi snapshot), getting and paying the models, and so on. If you compare this to prompt-based AI, some complex prompts may be judged as creative work, others are too simple to count as it. Changing the color, contrast and sharpness of an existing image also doesn't necessarily count as original artwork (but see some of Andy Warhol's work for differing opinions, of course).
Is it a licence (in practice) for a right to brand name or really the look? Could Forza make a model that looks almost exactly like a Ferrari but name it Furrari?
You know you're comparing the best example of photography with the worst example of using an AI, right?
A photographer may carefully study their craft and set up their shots. They can also be someone who puts no thought at all into taking an almost random photo. Same as someone can carefully tune the parameters of a model and refine their prompt until they get something which meets their exacting specifications, or they can be someone who grunts into a text box and clicks generate.
Playing around with Stable Diffusion running locally, comparing my output to the things held up as AI art removes any doubt in my mind as to whether the creators are artists. They are.