Copyright holders have been attacking copyright-infringing tools for quite a while now, there's nothing new about that. Imagine an image-generating ML dataset

Copyright holders have been attacking copyright-infringing tools for quite a while now, there's nothing new about that. Imagine an image-generating ML dataset that has been trained on nothing but images of Mickey Mouse. Surely the Disney corporation would come after that, I have no doubt about that. Why does commingling Mickey with other inputs change that, if the tool can still reproduce images of Mickey Mouse when asked for that?