Disagree. Authority is given Congress to establish an IP regime for the purpose of "promot[ing] the progress of science and useful arts". You would have to justify how banning gen AI is a. feasible at all, particularly with open-weight models; and b. how it "promotes the progress of useful arts." You would lose in court because it's very difficult to argue that keeping art as a skilled craftsman's trade is worse for its progress than lowering the barriers to individuals expressing what they see.
I think bad AI makes bad output and so a few people are worried it will replace good human art with bad AI art. Realistically, the stuff it's replacing now is bad human art: stock photos and clipart stuff that weren't really creative expression to start with. As it improves, we'll be increasingly able to go do a targeted inpaint to create images that more closely match our creative vision. There's a path here that lowers the barriers for someone getting his ideas into a visual form and that's an unambiguous good, unless you're one of the "craftsmen" who invested time to learn the old way.
It's almost exactly the same as AI development. As an experienced dev who knows the ins and outs really well I look at AI code and say, "wow, that's garbage." But people are using it to make unimportant webshit frontends, not do "serious work". Once it can do "serious work" that will decrease the number of jobs in the field but be good for software development as a whole.