It could be argued that the training of an AI model does not fulfil all of these requirements,but it is likely to be something that will be argued in court by future defendants incopyright litigation. It is clear that making a temporary copy for training is transient, butit could be said that it is not incidental. The copy is also part of a technological process,but it may not be considered a lawful use. Similarly, one could argue that the resultingmodel does have economic significance, but specific copies do not, a model such as StableDiffusion can be trained with billions of images, each individual copy used in training maynot count as having “independent economic significance”.