The Supreme Court’s recent decision in Andy Warhol Foundation for the Visual Arts, Inc. v. Goldsmith clarifies the scope of transformative use and the role of these uses in the fair use analysis. This important case has implications for a fair use analysis of artificial intelligence. This article evaluates the interaction between copyright law’s fair use doctrine and typical sources and uses for artificial intelligence. In other words, the article will assess whether or not the use of copyrighted material to “train” AI programs—AI inputs—and the products of AI programs—AI outputs—are likely to found to be transformative in light of the Warhol framework. This article assesses the potential fair use analysis for generative AI applications in light of Warhol’s analytical framework. The central question in Warhol is the scope of transformative use versus a use that is derivative and which supplants a market for the original copyrighted work. Whether the use of copyrighted material to “train” AI programs and the products of AI programs are likely to found to be transformative in light of the Warhol framework is an intensely factual inquiry. This article concludes that the use of copyrighted material as inputs for training AI programs is — by itself—likely to be found to be a transformative fair use in most circumstances. The more difficult question is how AI outputs are analyzed. Fair use is necessarily a case-by-case inquiry. In light of cases like Warhol and Google v. Oracle, the analysis will turn on a series of considerations that are identified in this article. It is likely that the fair use question will be litigated frequently in the context of AI outputs, which can involve myriad factual scenarios.